Finite size effects on the phase diagram of a binary mixture confined between competing wal
RFC1242
Internet RFC/STD/FYI/BCP ArchivesRFC1242[ Index | Search | What's New | Comments | Help ]Network Working Group S. Bradner, Editor Request for Comments: 1242 Harvard University July 1991Benchmarking Terminology for Network Interconnection DevicesStatus of this MemoThis memo provides information for the Internet community. It does not specify an Internet standard. Distribution of this memo isunlimited.AbstractThis memo discusses and defines a number of terms that are used indescribing performance benchmarking tests and the results of suchtests. The terms defined in this memo will be used in additionalmemos to define specific benchmarking tests and the suggested format to be used in reporting the results of each of the tests. This memo is a product of the Benchmarking Methodology Working Group (BMWG) of the Internet Engineering Task Force (IETF).1. IntroductionVendors often engage in "specsmanship" in an attempt to give theirproducts a better position in the marketplace. This usually involves much "smoke & mirrors" used to confuse the user. This memo andfollow-up memos attempt to define a specific set of terminology and tests that vendors can use to measure and report the performancecharacteristics of network devices. This will provide the usercomparable data from different vendors with which to evaluate these devices.2. Definition formatTerm to be defined. (e.g., Latency)Definition:The specific definition for the term.Discussion:A brief discussion about the term, it's applicationand any restrictions on measurement procedures.Measurement units:The units used to report measurements of thisterm, if applicable.Issues:List of issues or conditions that effect this term.See Also:List of other terms that are relevant to the discussion of this term.3. Term definitions3.1 Back-to-backDefinition:Fixed length frames presented at a rate such that there is the minimum legal separation for a given mediumbetween frames over a short to medium period of time,starting from an idle state.Discussion:A growing number of devices on a network can producebursts of back-to-back frames. Remote disk serversusing protocols like NFS, remote disk backup systemslike rdump, and remote tape access systems can beconfigured such that a single request can result ina block of data being returned of as much as 64K octets. Over networks like ethernet with a relatively small MTU this results in many fragments to be transmitted. Since fragment reassembly will only be attempted if allfragments have been received, the loss of even onefragment because of the failure of some intermediatenetwork device to process enough continuous frames can cause an endless loop as the sender repetitivelyattempts to send its large data block.With the increasing size of the Internet, routingupdates can span many frames, with modern routers able to transmit very quickly. Missing frames of routinginformation can produce false indications ofunreachability. Tests of this parameter are intendedto determine the extent of data buffering in thedevice.Measurement units:Number of N-octet frames in burst.Issues:See Also:3.2 BridgeDefinition:A system which forwards data frames based on information in the data link layer.Discussion:Measurement units:n/aIssues:See Also:bridge/router (3.3)router (3.15)3.3 bridge/routerDefinition:A bridge/router is a network device that can selectively function as a router and/or a bridge based on theprotocol of a specific frame.Discussion:Measurement units:n/aIssues:See Also:bridge (3.2)router (3.15)3.4 Constant LoadDefinition:Fixed length frames at a fixed interval time.Discussion:Although it is rare, to say the least, to encountera steady state load on a network device in the realworld, measurement of steady state performance maybe useful in evaluating competing devices. Theframe size is specified and constant. All deviceparameters are constant. When there is a checksumin the frame, it must be verified.Measurement units:n/aIssues:unidirectional vs. bidirectionalSee Also:3.5 Data link frame sizeDefinition:The number of octets in the frame from the first octet following the preamble to the end of the FCS, ifpresent, or to the last octet of the data if thereis no FCS.Discussion:There is much confusion in reporting the framesizes used in testing network devices or networkmeasurement. Some authors include the checksum,some do not. This is a specific definition for usein this and subsequent memos.Measurement units:octetsIssues:See Also:3.6 Frame Loss RateDefinition:Percentage of frames that should have been forwardedby a network device under steady state (constant)load that were not forwarded due to lack ofresources.Discussion:This measurement can be used in reporting theperformance of a network device in an overloadedstate. This can be a useful indication of how adevice would perform under pathological networkconditions such as broadcast storms.Measurement units:Percentage of N-octet offered frames that are dropped. To be reported as a graph of offered load vs frame loss.Issues:See Also:overhead behavior (3.11)policy based filtering (3.13)MTU mismatch behavior (3.10)3.7 Inter Frame GapDefinition:The delay from the end of a data link frame as defined in section 3.5, to the start of the preamble of thenext data link frame.Discussion:There is much confusion in reporting the betweenframe time used in testing network devices. Thisis a specific definition for use in this and subsequent memos.Measurement units:Time with fine enough units to distinguish between2 events.Issues:Link data rate.See Also:3.8 LatencyDefinition:For store and forward devices:The time interval starting when the last bit of theinput frame reaches the input port and ending whenthe first bit of the output frame is seen on theoutput port.For bit forwarding devices:The time interval starting when the end of the firstbit of the input frame reaches the input port andending when the start of the first bit of the outputframe is seen on the output port.Discussion:Variability of latency can be a problem.Some protocols are timing dependent (e.g., LAT and IPX). Future applications are likely to be sensitive tonetwork latency. Increased device delay can reducethe useful diameter of net. It is desired toeliminate the effect of the data rate on the latencymeasurement. This measurement should only reflect the actual within device latency. Measurements should betaken for a spectrum of frame sizes without changingthe device setup.Ideally, the measurements for all devices would be fromthe first actual bit of the frame after the preamble. Theoretically a vendor could design a device thatnormally would be considered a store and forwarddevice, a bridge for example, that begins transmitting a frame before it is fully received. This type ofdevice is known as a "cut through" device. Theassumption is that the device would somehow invalidate the partially transmitted frame if in receiving theremainder of the input frame, something came up that the frame or this specific forwarding of it was inerror. For example, a bad checksum. In this case,the device would still be considered a store andforward device and the latency would still befrom last bit in to first bit out, even though thevalue would be negative. The intent is to treatthe device as a unit without regard to the internalstructure.Measurement units:Time with fine enough units to distinguish between2 events.Issues:See Also:link speed mismatch (3.9)constant load (3.4)back-to-back (3.1)policy based filtering (3.13)single frame behavior (3.16)3.9 Link Speed MismatchDefinition:Speed mismatch between input and output data rates.Discussion:This does not refer to frame rate per se, it refers to the actual data rate of the data path. For example,an Ethernet on one side and a 56KB serial link on the other. This is has also been referred to as the "fire hose effect". Networks that make use of serial links between local high speed networks will usually havelink speed mismatch at each end of the serial links.Measurement units:Ratio of input and output data rates.Issues:See Also:constant load (3.4)back-to-back (3.1)3.10 MTU-mismatch behaviorDefinition:The network MTU (Maximum Transmission Unit) of theoutput network is smaller than the MTU of the inputnetwork, this results in fragmentation.Discussion:The performance of network devices can be significantly affected by having to fragment frames.Measurement units:Description of behavior.Issues:See Also:3.11 Overhead behaviorDefinition:Processing done other than that for normal data frames.Discussion:Network devices perform many functions in additionto forwarding frames. These tasks range from internal hardware testing to the processing of routinginformation and responding to network managementrequests. It is useful to know what the effect ofthese sorts of tasks is on the device performance.An example would be if a router were to suspendforwarding or accepting frames during the processingof large routing update for a complex protocol likeOSPF. It would be good to know of this sort ofbehavior.Measurement units:Any quantitative understanding of this behavior is by the determination of its effect on other measurements.Issues:bridging and routing protocolscontrol processingicmpip options processingfragmentationerror processingevent logging/statistics collectionarpSee Also:policy based filtering (3.13)3.12 Overloaded behaviorDefinition:When demand exceeds available system resources.Discussion:Devices in an overloaded state will lose frames. The device might lose frames that contain routing orconfiguration information. An overloaded state isassumed when there is any frame loss.Measurement units:Description of behavior of device in any overloadedstates for both input and output overload conditions.Issues:How well does the device recover from overloaded state? How does source quench production effect device?What does device do when its resources are exhausted? What is response to system management in overloadedstate?See Also:3.13 Policy based filteringDefinition:Filtering is the process of discarding receivedframes by administrative decision where normaloperation would be to forward them.Discussion:Many network devices have the ability to beconfigured to discard frames based on a numberof criteria. These criteria can range from simplesource or destination addresses to examiningspecific fields in the data frame itself.Configuring many network devices to performfiltering operations impacts the throughputof the device.Measurement units:n/aIssues:flexibility of filter optionsnumber of filter conditionsSee Also:3.14 Restart behaviorDefinition:Reinitialization of system causing data loss.Discussion:During a period of time after a power up orreset, network devices do not accept and forwardframes. The duration of this period of unavailabilitycan be useful in evaluating devices. In addition,some network devices require some form of resetwhen specific setup variables are modified. If thereset period were long it might discourage networkmanagers from modifying these variables on productionnetworks.Measurement units:Description of device behavior under various restartconditions.Issues:Types:power onreload software imageflush port, reset buffersrestart current code image, without reconfurationUnder what conditions is a restart required?Does the device know when restart needed (i.e., hungstate timeout)?Does the device recognize condition of too frequentauto-restart?Does the device run diagnostics on all or some resets?How may restart be initiated?physical interventionremote via terminal line or login over networkSee Also:3.15 RouterDefinition:A system which forwards data frames based oninformation in the network layer.Discussion:This implies "running" the network level protocolrouting algorithm and performing whatever actionsthat the protocol requires. For example, decrementingthe TTL field in the TCP/IP header.Measurement units:n/aIssues:See Also:bridge (3.2)bridge/router (3.3)3.16 Single frame behaviorDefinition:One frame received on the input to a device.Discussion:A data "stream" consisting of a single frame canrequire a network device to do a lot of processing.Figuring routes, performing ARPs, checkingpermissions etc., in general, setting up cache entries. Devices will often take much more time to process asingle frame presented in isolation than it would ifthe same frame were part of a steady stream. Thereis a worry that some devices would even discard a single frame as part of the cache setup procedure under theassumption that the frame is only the first of many.Measurement units:Description of the behavior of the device.Issues:See Also:policy based filtering (3.13)3.17 ThroughputDefinition:The maximum rate at which none of the offered framesare dropped by the device.Discussion:The throughput figure allows vendors to report asingle value which has proven to have use in themarketplace. Since even the loss of one frame in adata stream can cause significant delays whilewaiting for the higher level protocols to time out,it is useful to know the actual maximum datarate that the device can support. Measurements should be taken over a assortment of frame sizes. Separatemeasurements for routed and bridged data in thosedevices that can support both. If there is a checksum in the received frame, full checksum processing mustbe done.Measurement units:N-octet input frames per secondinput bits per secondIssues:single path vs. aggregateloadunidirectional vs bidirectionalchecksum processing required on some protocolsSee Also:frame loss rate (3.6)constant load (3.4)back-to-back (3.1)4. AcknowledgementsThis memo is a product of the IETF BMWG working group:Chet Birger, Coral NetworksScott Bradner, Harvard University (chair)Steve Butterfield, independant consultantFrank Chui, TRWPhill Gross, CNRIStev Knowles, FTP Software, Inc.Mat Lew, TRWGary Malkin, FTP Software, Inc.K.K. Ramakrishnan, Digital Equipment Corp.Mick Scully, Ungerman BassWilliam M. Seifert, Wellfleet Communications Corp.John Shriver, Proteon, Inc.Dick Sterry, MicrocomGeof Stone, Network Systems Corp.Geoff Thompson, SynOpticsMary Youssef, IBMSecurity ConsiderationsSecurity issues are not discussed in this memo.Author's AddressScott BradnerHarvard UniversityWilliam James Hall 123233 Kirkland StreetCambridge, MA 02138Phone: (617) 495-3864EMail: SOB@Or, send comments to: bmwg@.[ Index | Search | What's New | Comments | Help ] Comments/Questions about this archive ? Send mail to rfc-admin@。
手征极限下带温度和化学势的两味道Wilson费米子QCD的相解读
The Phase diagram suggested by Roberge and Weiss
First order
Some
observables
considered
Nt 1 t 0
Polyakov loop Chiral condensate
P( x ) Tr[U t ( x )]
with β=6/g2
利用 Wilson 费米子, 则费米子矩阵为:
在需要考虑化学势时,代换费米子作用量中时间方向的链, 引入化学势。
但是:
引入化学势后, 对SU(3)
M ( 5† M 5 )† , ( 0) M ( 5† M 5 )† , ( 0)
费米子矩阵的行列式为复数, 使得Monte Carlo模拟不能进行。
N inv
From hot start cold start
Results from the scanning along the temperature axis, i.e. beta axis.
Critical beta as a function of imaginary chemical potential
• 连续的夸克作用量
• 在格点上代换为离散的夸克作用量
M
是离散的费米子矩阵
解决办法
a. Improved reweighting
b. Imaginary chemical potential
III.
Lattice QCD with Imaginary Chemical Potential With Wilson Quarks
II. Lattice Formulation
N f 味夸克的系统的配分函数为(带有化学势)
一些常见的统计术语翻译
一些常见的统计术语翻译Absolute deviation, 绝对离差Absolute number , 绝对数Absolute r esiduals, 绝对残差Acceler ation arr ay, 加速度立体阵Acceler ation in an arbitr ary dir ection, 任意方向上的加速度Acceler ation nor mal, 法向加速度Acceler ation spac e dimension, 加速度空间的维数Acceler ation tangential, 切向加速度Acceler ation vector , 加速度向量Acceptable hypothesis, 可接受假设Accum ulation, 累积Accuracy, 准确度Actual fr equency, 实际频数Adaptive estimator , 自适应估计量Addition, 相加Addition theor em , 加法定理Additivity, 可加性Adjusted r ate, 调整率Adjusted value, 校正值Adm issible error , 容许误差Aggregation, 聚集性Alternative hypothesis, 备择假设Among gr oups, 组间Amounts, 总量Analysis of c orr elation, 相关分析Analysis of c ovarianc e, 协方差分析Analysis of r egr ession, 回归分析Analysis of time series, 时间序列分析Analysis of varianc e, 方差分析Angular tr ansfor mation, 角转换ANOVA (analysis of variance ), 方差分析ANOVA Models, 方差分析模型Arcing, 弧/ 弧旋Arcsine tr ansfor mation, 反正弦变换Area under the curve, 曲线面积AREG , 评估从一个时间点到下一个时间点回归相关时的误差ARIMA, 季节和非季节性单变量模型的极大似然估计Arithmetic grid paper , 算术格纸Arithmetic mean, 算术平均数Arrhenius r elation, 艾恩尼斯关系Assessing fit, 拟合的评估Associative laws, 结合律Asymmetric distribution, 非对称分布Asymptotic bias, 渐近偏倚Asymptotic efficiency, 渐近效率Asymptotic variance, 渐近方差Attributable risk, 归因危险度Attribute data, 属性资料Attribution, 属性Autoc orrelation, 自相关Autoc orrelation of residuals, 残差的自相关Aver age, 平均数Aver age c onfidenc e interval length, 平均置信区间长度Aver age growth r ate, 平均增长率Bar c hart, 条形图Bar gr aph, 条形图Base period, 基期Bayes' theorem , Bayes 定理Bell-shaped curve, 钟形曲线伯努力分布Ber noulli distribution,Best-trim estimator , 最好切尾估计量Bias, 偏性Binary logistic r egr ession, 二元逻辑斯蒂回归Binomial distribution, 二项分布Bisquare, 双平方Bivariate Corr elate, 二变量相关Bivariate nor mal distribution, 双变量正态分布Bivariate nor mal population, 双变量正态总体Biweight inter val, 双权区间Biweight M-estimator, 双权M 估计量Bloc k, 区组/ 配伍组BMDP(Biomedic al computer pr ograms), BMDP 统计软件包Boxplots, 箱线图/ 箱尾图Breakdown bound, 崩溃界/ 崩溃点Canonical c orrelation, 典型相关Caption, 纵标目Case-c ontrol study , 病例对照研究Categoric al variable, 分类变量Catenary, 悬链线Cauchy distribution, 柯西分布Cause-and-effect r elationship, 因果关系Cell, 单元Censoring, 终检Center of symmetry , 对称中心Centering and sc aling, 中心化和定标Centr al tendency, 集中趋势Centr al value, 中心值CHAID - x 2 Automatic Inter action Detector ,卡方自动交互检测Chanc e, 机遇Chanc e error , 随机误差Chanc e variable, 随机变量Char acteristic equation, 特征方程Char acteristic root, 特征根Char acteristic vector , 特征向量Chebshev criterion of fit, 拟合的切比雪夫准则Chernoff fac es, 切尔诺夫脸谱图Chi-square test, 卡方检验/咒2检验Choleskey dec omposition, 乔洛斯基分解Circle chart, 圆图Class interval, 组距Class mid-value, 组中值Class upper limit, 组上限Classified variable, 分类变量Cluster analysis, 聚类分析Cluster sampling, 整群抽样Code, 代码Coded data, 编码数据Coding, 编码Coefficient of c ontingency, 列联系数Coefficient of deter mination, 决定系数Coefficient of multiple c orr elation, 多重相关系数Coefficient of partial c orrelation, 偏相关系数Coefficient of pr oduction-moment c orrelation, 积差相关系数Coefficient of r ank corr elation, 等级相关系数Coefficient of r egr ession, 回归系数Coefficient of skewness, 偏度系数Coefficient of variation, 变异系数Cohort study, 队列研究Column, 列Column effect, 列效应Column factor , 列因素Combination pool, 合并Combinative table, 组合表Common factor , 共性因子Common regr ession coefficient, 公共回归系数Common value, 共同值Common varianc e, 公共方差Common variation, 公共变异Communality varianc e, 共性方差Compar ability, 可比性Comparison of bathes, 批比较Comparison value, 比较值Compartment model, 分部模型Compassion, 伸缩Complement of an event, 补事件Complete association, 完全正相关Complete dissociation, 完全不相关Complete statistic s, 完备统计量Completely r andomized design, 完全随机化设计Composite event, 联合事件Composite events, 复合事件Concavity, 凹性Conditional expectation, 条件期望Conditional likelihood, 条件似然Conditional pr obability, 条件概率Conditionally linear , 依条件线性Confidenc e interval, 置信区间Confidenc e lim it, 置信限Confidenc e lower lim it, 置信下限Confidenc e upper limit, 置信上限Confir matory Factor Analysis , 验证性因子分析Confir matory research, 证实性实验研究Confounding factor , 混杂因素Conjoint, 联合分析Consistency, 相合性Consistency chec k, 一致性检验Consistent asymptotic ally nor mal estimate, 相合渐近正态估计Consistent estimate, 相合估计Constr ained nonlinear r egr ession, 受约束非线性回归Constr aint, 约束Contam inated distribution, 污染分布Contam inated Gausssian, 污染高斯分布Contam inated nor mal distribution, 污染正态分布Contam ination, 污染Contam ination model, 污染模型Contingency table, 列联表Contour , 边界线Contribution r ate, 贡献率Control, 对照Controlled experiments, 对照实验Conventional depth, 常规深度Convolution, 卷积Corrected factor , 校正因子Corrected mean, 校正均值Correction coefficient, 校正系数Correctness, 正确性Correlation c oefficient, 相关系数Correlation index, 相关指数Correspondenc e, 对应Counting, 计数Counts, 计数/ 频数Covarianc e, 协方差Covariant, 共变Cox Regression, Cox 回归Criteria for fitting, 拟合准则Criteria of least squar es, 最小二乘准则Critic al r atio, 临界比Critic al r egion, 拒绝域Critic al value, 临界值Cr oss-over design, 交叉设计Cr oss-section analysis, 横断面分析Cr oss-section survey, 横断面调查Cr osstabs , 交叉表Cr oss-tabulation table, 复合表Cube r oot, 立方根Cumulative distribution function, 分布函数Cumulative probability, 累计概率Curvatur e, 曲率/ 弯曲Curvatur e, 曲率Curve fit , 曲线拟和Curve fitting, 曲线拟合Curvilinear r egression, 曲线回归Curvilinear r elation, 曲线关系Cut-and-try method, 尝试法Cycle, 周期Cyclist, 周期性D test, D 检验Data acquisition, 资料收集Data bank, 数据库Data c apacity, 数据容量Data deficiencies, 数据缺乏Data handling, 数据处理Data manipulation, 数据处理Data proc essing, 数据处理Data r eduction, 数据缩减Data set, 数据集Data sourc es, 数据来源Data tr ansfor mation, 数据变换Data validity, 数据有效性Data-in, 数据输入Data-out, 数据输出Dead time, 停滞期Degr ee of fr eedom, 自由度Degr ee of pr ecision, 精密度Degr ee of r eliability , 可靠性程度Degr ession, 递减Density function, 密度函数Density of data points,数据点的密度Dependent variable,应变量/ 依变量/ 因变量Dependent variable,因变量Depth, 深度Derivative matrix, 导数矩阵Derivative-fr ee methods, 无导数方法Design, 设计Deter minacy, 确定性Deter minant, 行列式Deter minant, 决定因素Deviation, 离差Deviation from aver age, 离均差Diagnostic plot, 诊断图Dichotomous variable, 二分变量Differential equation,微分方程Direct standardization, 直接标准化法Discr ete variable, 离散型变量DISCRIMINAN T, 判断Discriminant analysis, 判别分析Discriminant c oeffic ient, 判别系数Discriminant function, 判别值Disper sion, 散布/ 分散度Dispr oportional, 不成比例的Dispr oportionate sub-class numbers, 不成比例次级组含量Distribution free, 分布无关性/ 免分布Distribution shape, 分布形状Distribution-free method, 任意分布法Distributive laws, 分配律Distur banc e, 随机扰动项Dose response curve, 剂量反应曲线Double blind method, 双盲法Double blind trial, 双盲试验Double exponential distribution, 双指数分布Double logarithmic, 双对数Downward r ank, 降秩Dual-spac e plot, 对偶空间图DUD, 无导数方法新法Duncan's new multiple r ange method, 新复极差法/DuncanE-LEffect, 实验效应Eigenvalue, 特征值Eigenvector , 特征向量Ellipse, 椭圆Empiric al distribution, 经验分布Empiric al pr obability , 经验概率单位Enumer ation data, 计数资料Equal sun-class number , 相等次级组含量Equally likely , 等可能Equivarianc e, 同变性Error , 误差/ 错误Errorof estimate, 估计误差Error type I, 第一类错误Error type II, 第二类错误Estimand, 被估量Estimated err or mean squares, 估计误差均方Estimated err or sum of squar es, 估计误差平方和Euclidean distanc e,欧式距离Event, 事件Event, 事件Exc eptional data point, 异常数据点Expectation plane, 期望平面Expectation surfac e, 期望曲面Expected values, 期望值Experiment, 实验Experimental sampling, 试验抽样Experimental unit, 试验单位Explanatory variable, 说明变量Explor atory data analysis, 探索性数据分析Explore Summarize, 探索- 摘要Exponential curve, 指数曲线Exponential growth, 指数式增长EXSMOOTH, 指数平滑方法Extended fit, 扩充拟合Extr a par ameter ,附加参数Extr apolation, 外推法Extr eme observation, 末端观测值Extr emes, 极端值/ 极值F distribution, F分布 F test, F 检验Factor , 因素 / 因子Factor analysis, 因子分析Factor Analysis, 因子分析Factor scor e, 因子得分Factorial, 阶乘Factorial design, 析因试验设计False negative, 假阴性False negative error , 假阴性错误 Fam ily of distributions, 分布族 Fam ily of estimator s, 估计量族 Fanning, 扇面Fatality r ate, 病死率Field investigation, 现场调查Field survey , 现场调查Finite population, 有限总体 Finite-sample, 有限样本First derivative, 一阶导数First principal component,First quartile, 第一四分位数Fisher infor mation, 费雪信息量Fitted value, 拟合值Fourth, 四分点Frequency, 频率Frontier point, 界限点Function r elationship, 泛函关系Gaussian distribution, 高斯分布 / 正态分布Gini's mean difference,基尼均差 GLM (Gener al liner models), 通用线性模型Fitting a c urve, 曲线拟合 Fixed base,定基 Fluctuation, 随机起伏 For ec ast, 预测 Four fold table,四格表Fraction blow, 左侧比率Fractional error, 相对误差 Frequency polygon,频数多边图 Gamma distribution, 伽玛分布Gauss increment, 高斯增量Gauss-Newton incr ement, 高斯- 牛顿增量 Gener al census, 全面普查GENLOG (Gener alized liner models), 广义线性模型 Geometric mean,几何平均数 第一主成分Goodness of fit, 拟和优度/ 配合度Gradient of deter m inant, 行列式的梯度Graec o-Latin squar e, 希腊拉丁方Grand mean, 总均值Gross error s, 重大错误Gross-error sensitivity, 大错敏感度Group aver ages, 分组平均Grouped data, 分组资料Guessed mean, 假定平均数Half-life, 半衰期Hampel M-estimators, 汉佩尔M 估计量Happenstanc e, 偶然事件Har monic mean, 调和均数Hazar d function, 风险均数Hazar d r ate, 风险率Heading, 标目Heavy-tailed distribution, 重尾分布Hessian arr ay, 海森立体阵Heterogeneity , 不同质Heterogeneity of variance, 方差不齐Hier archic al classific ation, 组内分组Hier archic al clustering method, 系统聚类法High-lever age point, 高杠杆率点HILOGLINEAR, 多维列联表的层次对数线性模型Hinge, 折叶点Histogr am, 直方图Historical c ohort study, 历史性队列研究Holes, 空洞HOMALS, 多重响应分析Homogeneity of varianc e, 方差齐性Homogeneity test, 齐性检验Huber M-estimators, 休伯M 估计量Hyper bola, 双曲线Hypothesis testing, 假设检验Hypothetic al universe, 假设总体Impossible event, 不可能事件Independenc e, 独立性Independent variable, 自变量Index, 指标/ 指数Indir ect standardization, 间接标准化法Individual, 个体Infer enc e band, 推断带Infinite population, 无限总体Infinitely gr eat, 无穷大Infinitely small, 无穷小Influence curve, 影响曲线Intercept, 截距Interpolation, 内插法Invarianc e, 不变性Inverse matrix, 逆矩阵Inverse sine tr ansfor mation, 反正弦变换Iter ation, 迭代Jac obian deter m inant, 雅可比行列式Joint distribution function,分布函数 Joint probability, 联合概率Joint probability distribution,联合概率分布 K means method, 逐步聚类法Kaplan-Meier , 评估事件的时间长度Kaplan-Merier c hart, Kaplan-Merier图 Kendall's r ank c orrelation, Kendall等级相关 Kinetic, 动力学Kolmogor ov-Smirnove test, 柯尔莫哥洛夫 - 斯米尔诺夫检验Kruskal and Wallis test, Kr uskal 及 Wallis 检验 / 多样本的秩和检验 /H 检验 Kurtosis, 峰度Lac k of fit, 失拟Ladder of powers, 幂阶梯Lag, 滞后Lar ge sample, 大样本Lar ge sample test, 大样本检验Latin squar e, 拉丁方Latin squar e design, 拉丁方设计Leakage, 泄漏Least favor able c onfigur ation, 最不利构形Least favor able distribution, 最不利分布Least signific ant differ enc e, 最小显著差法Least squar e method, 最小二乘法Least-absolute-r esiduals estimates, Least-absolute-r esiduals fit, 最小绝对残差拟合 Least-absolute-r esiduals line, 最小绝对残差线 Legend, 图例L-estimator , L 估计量Infor mation capacity, 信息容量 Initial condition,初始条件 Initial estimate,初始估计值 Initial level,最初水平 Interaction,交互作用 Interaction terms, 交互作用项Interquartile range,四分位距 Interval estimation,区间估计 Intervals of equal probability, 等概率区间 Intrinsic c urvature,固有曲率Inverse probability,逆概率最小绝对残差估计L-estimator of loc ation, 位置L 估计量L-estimator of sc ale, 尺度L 估计量Level, 水平Life expectanc e, 预期期望寿命Life table, 寿命表Life table method, 生命表法Light-tailed distribution, 轻尾分布似然函数Likelihood function,似然比Likelihood r atio,line gr aph, 线图直线相关Linear corr elation,线性方程Linear equation,Linear pr ogr amm ing, 线性规划直线回归Linear regr ession,线性回归Linear Regression,Linear trend, 线性趋势Loading, 载荷Loc ation and sc ale equivarianc e, 位置尺度同变性Loc ation equivarianc e, 位置同变性Loc ation invarianc e, 位置不变性Loc ation sc ale family, 位置尺度族Log r ank test, 时序检验Logarithm ic curve, 对数曲线Logarithm ic nor mal distribution, 对数正态分布Logarithm ic sc ale, 对数尺度Logarithm ic tr ansfor mation, 对数变换Logic chec k, 逻辑检查Logistic distribution, 逻辑斯特分布Logit tr ansfor mation, Logit 转换LOGLINEAR, 多维列联表通用模型Lognor mal distribution, 对数正态分布Lost function, 损失函数Low corr elation, 低度相关Lower lim it, 下限Lowest-attained varianc e, 最小可达方差LSD, 最小显著差法的简称Lur king variable, 潜在变量M-RMain effect, 主效应Major heading, 主辞标目Marginal density function, 边缘密度函数Marginal pr obability, 边缘概率Marginal pr obability distribution, 边缘概率分布Matched data, 配对资料Matched distribution, 匹配过分布Matching of distribution, 分布的匹配Matching of tr ansfor mation, 变换的匹配Mathematic al expectation, 数学期望Mathematic al model, 数学模型Maximum L-estimator , 极大极小L 估计量Maximum likelihood method, 最大似然法Mean, 均数Mean squar es between groups, 组间均方Mean squar es within gr oup, 组内均方Means (Compar e means), 均值- 均值比较Median, 中位数Median effective dose, 半数效量Median lethal dose, 半数致死量Median polish, 中位数平滑Median test, 中位数检验Minimal sufficient statistic, 最小充分统计量Minimum distanc e estimation, 最小距离估计Minimum effective dose, 最小有效量Minimum lethal dose, 最小致死量Minimum varianc e estimator , 最小方差估计量MIN ITAB, 统计软件包Minor heading, 宾词标目Missing data, 缺失值Model specific ation, 模型的确定Modeling Statistic s , 模型统计Models for outliers, 离群值模型Modifying the model, 模型的修正Modulus of c ontinuity , 连续性模Mor bidity , 发病率Most favor able c onfigur ation, 最有利构形Multidimensional Sc aling (ASCAL), 多维尺度/ 多维标度Multinomial Logistic Regression , 多项逻辑斯蒂回归Multiple c omparison, 多重比较Multiple c orr elation , 复相关Multiple c ovarianc e, 多元协方差Multiple linear r egr ession, 多元线性回归Multiple r esponse , 多重选项Multiple solutions, 多解Multiplic ation theor em , 乘法定理Multir esponse, 多元响应Multi-stage sampling, 多阶段抽样Multivariate T distribution, 多元T 分布Mutual exclusive, 互不相容Mutual independenc e, 互相独立Natur al boundary, 自然边界Natur al dead, 自然死亡Natur al zer o, 自然零Negative c orr elation, 负相关Negative linear corr elation, 负线性相关Negatively skew ed, 负偏Newman-Keuls method, q 检验NK method, q 检验No statistic al signific ance, 无统计意义Nom inal variable, 名义变量Nonc onstancy of variability, 变异的非定常性Nonlinear regr ession, 非线性相关Nonpar ametric statistics, 非参数统计Nonpar ametric test, 非参数检验Nonpar ametric tests, 非参数检验Normal deviate, 正态离差Normal distribution, 正态分布Normal equation, 正规方程组Normal r anges, 正常范围Normal value, 正常值Nuisanc e par ameter , 多余参数/ 讨厌参数Null hypothesis, 无效假设Numeric al variable, 数值变量Objective function, 目标函数观察单位Observation unit,观察值Observed value,One sided test, 单侧检验One-way analysis of varianc e, 单因素方差分析Oneway ANOVA , 单因素方差分析Open sequential trial, 开放型序贯设计Optrim, 优切尾Optrim efficiency, 优切尾效率Order statistic s, 顺序统计量Or dered categories, 有序分类Or dinal logistic r egr ession , 序数逻辑斯蒂回归有序变量Or dinal variable,正交基Orthogonal basis,Orthogonal design, 正交试验设计Orthogonality c onditions, 正交条件ORTHOPLAN, 正交设计Outlier cutoffs, 离群值截断点Outlier s, 极端值OVE RALS , 多组变量的非线性正规相关Over shoot, 迭代过度Pair ed design, 配对设计Pair ed sample, 配对样本Pairwise slopes, 成对斜率Par abola, 抛物线Par allel tests, 平行试验Par ameter , 参数Par ametric statistic s, 参数统计Par ametric test, 参数检验Partial c orrelation, 偏相关Partial r egression, 偏回归Partial sorting, 偏排序Partials r esiduals, 偏残差Patter n, 模式Pear son curves, 皮尔逊曲线Peeling, 退层Perc ent bar gr aph, 百分条形图Perc entage, 百分比Perc entile, 百分位数Perc entile curves, 百分位曲线Periodicity , 周期性Per mutation, 排列P-estimator , P 估计量Pie graph, 饼图Pitman estimator , 皮特曼估计量Pivot, 枢轴量Planar , 平坦Planar assumption, 平面的假设PLANCARDS, 生成试验的计划卡Point estimation, 点估计Poisson distribution, 泊松分布Polishing, 平滑Polled standar d deviation, 合并标准差Polled varianc e, 合并方差Polygon, 多边图Polynomial, 多项式Polynomial c urve, 多项式曲线Population, 总体Population attributable risk,人群归因危险度Qualitative classific ation, 属性分类Qualitative method, 定性方法Quantile-quantile plot, Quantitative analysis, Quartile, 四分位数Quic k Cluster , 快速聚类Radix sort, 基数排序Random alloc ation, 随机化分组Random bloc ks design, 随机区组设计Random event, 随机事件Random ization, 随机化Range, 极差/ 全距Rank c orr elation, 等级相关Rank sum test, 秩和检验Rank test, 秩检验 Ranked data, 等级资料Rate, 比率Ratio, 比例 Positive c orrelation, 正相关Positively skewed, 正偏Posterior distribution, 后验分布Power of a test, 检验效能 Precision,精密度Predicted value, 预测值Preliminary analysis, 预备性分析Principal c omponent analysis, 主成分分析Prior distribution, 先验分布 Prior pr obability, Probabilistic model, probability, 概率Probability density Product moment, 先验概率概率模型, 概率密度 乘积矩 / 协方差Profile tr ace, 截面迹图Proportion, 比/ 构成比Proportion alloc ation in str atified random sampling, Proportionate, 成比例Proportionate sub-class numbers, 成比例次级组含量Prospective study , 前瞻性调查Proximities, 亲近性Pseudo F test, 近似 F 检验Pseudo model, 近似模型Pseudosigma, 伪标准差Purposive sampling, 有目的抽样QR dec omposition, QR 分解Quadratic approximation, 二次近似 按比例分层随机抽样分位数-分位数图 /Q-Q 图 定量分析Raw data, 原始资料Raw residual, 原始残差Rayleigh's test, 雷氏检验Rayleigh's Z, 雷氏Z 值Recipr ocal, 倒数Recipr ocal tr ansfor mation, 倒数变换Rec or ding, 记录Redesc ending estimators, 回降估计量Reducing dimensions, 降维Re-expression, 重新表达Refer enc e set, 标准组Region of acc eptanc e, 接受域Regr ession coefficient, 回归系数Regr ession sum of squar e, 回归平方和Rej ection point, 拒绝点Relative disper sion, 相对离散度Relative number , 相对数Reliability , 可靠性Repar ametrization, 重新设置参数Replication, 重复Report Summar ies, 报告摘要Residual sum of squar e, 剩余平方和Resistanc e, 耐抗性Resistant line, 耐抗线Resistant technique, 耐抗技术R-estimator of location, 位置R 估计量R-estimator of sc ale, 尺度R 估计量Retr ospective study, 回顾性调查Ridge tr ace, 岭迹Ridit analysis, Ridit 分析Rotation, 旋转Rounding, 舍入Row, 行Row effects, 行效应Row factor , 行因素RXC table, RXC 表S-ZSample, 样本Sample r egression c oefficient, 样本回归系数Sample size, 样本量Sample standar d deviation, 样本标准差Sampling error , 抽样误差SAS(Statistical analysis system ), SAS Scale, 尺度/ 量表Scatter diagr am, 散点图统计软件包Schematic plot, 示意图/ 简图Scor e test, 计分检验Screening, 筛检SEASON, 季节分析Sec ond derivative, 二阶导数Sec ond principal c omponent, 第二主成分SEM (Structur al equation modeling), 结构化方程模型Semi-logarithm ic gr aph, 半对数图Semi-logarithm ic paper , 半对数格纸Sensitivity c urve, 敏感度曲线Sequential analysis,贯序分析Sequential data set, 顺序数据集Sequential design, 贯序设计Sequential method, 贯序法Sequential test, 贯序检验法Serial tests, 系列试验Short-c ut method, 简捷法Sigmoid curve, S形曲线Sign function, 正负号函数Sign test, 符号检验Signed r ank, 符号秩Signific anc e test, 显著性检验Signific ant figur e, 有效数字Sim ple cluster sampling, 简单整群抽样Sim ple c orrelation, 简单相关Sim ple r andom sampling, 简单随机抽样Sim ple r egr ession, 简单回归simple table, 简单表Sine estimator , 正弦估计量Single-valued estimate, 单值估计Singular matrix, 奇异矩阵Skewed distribution, 偏斜分布Skewness, 偏度Slash distribution, 斜线分布Slope, 斜率Smirnov test, 斯米尔诺夫检验Source of variation, 变异来源Spear man r ank c orrelation, 斯皮尔曼等级相关Specific factor , 特殊因子Specific factor varianc e, 特殊因子方差Spectr a , 频谱Spherical distribution, 球型正态分布Spr ead, 展布SPSS(Statistical pac kage for the social scienc e), SPSS Spurious c orr elation, 假性相关Square root tr ansfor mation, 平方根变换Stabilizing variance, 稳定方差Standard deviation, 标准差Standard error , 标准误Standard error of differ ence, 差别的标准误Standard error of estimate, 标准估计误差Standard error of r ate, 率的标准误Standard nor mal distribution, 标准正态分布Standardization, 标准化Starting value, 起始值Statistic, 统计量Statistical c ontrol, 统计控制Statistical gr aph, 统计图Statistical inferenc e, 统计推断Statistical table, 统计表Steepest desc ent, 最速下降法Stem and leaf display, 茎叶图Step factor , 步长因子Stepwise r egr ession, 逐步回归Stor age, 存Strata, 层(复数)Stratified sampling, 分层抽样Stratified sampling, 分层抽样Strength, 强度Stringency , 严密性Structur al r elationship, 结构关系Studentized r esidual, 学生化残差/t 化残差Sub-class number s, 次级组含量Subdividing, 分割Sufficient statistic, 充分统计量Sum of pr oducts, 积和Sum of squares, 离差平方和Sur e event, 必然事件Survey, 调查Survival, 生存分析统计软件包Sum of squares about regr Sum of squares between gr Sum of squares of partial r ession, 回归平方和oups, 组间平方和egression, 偏回归平方和Survival r ate, 生存率Suspended r oot gr am, 悬吊根图Symmetry, 对称Systematic err or, 系统误差Systematic sampling, 系统抽样Tags, 标签Tail ar ea, 尾部面积Tail length, 尾长Tail weight, 尾重Tangent line, 切线Target distribution, 目标分布Taylor series, 泰勒级数Tendency of dispersion, 离散趋势Testing of hypotheses, 假设检验Theor etical frequency , 理论频数Time series, 时间序列Toler anc e interval, 容忍区间Toler anc e lower lim it, 容忍下限Toler anc e upper lim it, 容忍上限Torsion, 扰率Total sum of squar e, 总平方和Total variation, 总变异Transfor mation, 转换Treatment, 处理Trend, 趋势Trend of perc entage, 百分比趋势Trial, 试验Trial and err or method, 试错法Tuning c onstant, 细调常数Two sided test, 双向检验Two-stage least squar es, 二阶最小平方Two-stage sampling, 二阶段抽样Two-tailed test, 双侧检验Two-way analysis of varianc e, 双因素方差分析Two-way table, 双向表Type I err or, 一类错误/ a错误Type II err or,二类错误/ B错误UMVU, 方差一致最小无偏估计简称Unbiased estimate, 无偏估计Unc onstrained nonlinear r egr ession , 无约束非线性回归Unequal subclass number , 不等次级组含量Ungr ouped data, 不分组资料Unifor m coor dinate, 均匀坐标Unifor m distribution, 均匀分布Unifor m ly m inimum varianc e unbiased estimate, 方差一致最小无偏估计Unit, 单元Unor der ed categories, 无序分类Upper lim it, 上限Upwar d r ank, 升秩Vague conc ept, 模糊概念Validity , 有效性W test, W 检验W-estimation, W 估计量W-estimation of location,位置 W 估计量Width, 宽度 Wilcoxon paired test, 威斯康星配对法 / 配对符号秩和检验 Wild point, 野点 / 狂点Wild value, 野值 / 狂值Winsorized mean, 缩尾均值Withdr aw, 失访Youden's index, 尤登指数Z test, Z 检验Zer o corr elation, 零相关Z-tr ansfor mation, Z 变换 VARCOMP (Varianc e c omponent estimation), 方差元素估计 Variability , 变异性 Variable,变量 Varianc e,方差 Variation, 变异Varimax orthogonal rotation, 方差最大正交旋转 Volume of distribution,容积Weibull distribution, 威布尔分布 Weight, 权数Weighted Chi-squar e test, 加权卡方检验 /Coc hr an 检验 Weighted linear regression method, 加权直线回归 Weighted mean, 加权平均数Weighted mean squar Weighted sum of squarWeighting coefficient,Weighting method,e, 加权平均方差e, 加权平方和 权重系数 加权法。
UItrascale
Device Package User Guide UG112 (v3.7) September 5, 2012Chapter4Package Electrical CharacteristicsIntroductionAs data rates increase and signal rise times become shorter, the effects of package parasiticsare becoming increasingly significant as the hardware engineers model their circuits.Discontinuities that might have had minimal impact on circuit performance in pastgenerations of components are now of paramount importance as designers strive toachieve higher performance in their systems.The IC package forms an interconnect system just like traces on a printed circuit board(PCB) or conductors in connectors. When a designer simulates the signaling performancefrom a driver to a receiver, all the interconnect parasitics in the path, including the package,must be considered in order to achieve simulation results that represent the entire system'sperformance.Current Xilinx packages are constructed with either wirebond or flip chip interconnecttechnology. Some components use simpler leadframe-based packages, while others uselaminate-based packages with multilayer construction. The choice of package matches theperformance and marketing objectives sought for the device family. In multilayerpackages, innovative pin-out selections and creative design techniques are used in a co-design effort to optimize package performance and to prevent the package from being alimiting factor for the device. For these high performance FPGA packages, Xilinx alsoprovides package models that allow the user to take package parasites into account toaccurately model the component's performance prior to committing to hardware.This chapter focuses on defining certain critical concepts associated with electricalcharacterization of packages. It is also intended to provide relevant theoretical review ofelectrical issues and concepts as they relate to the characterization effort. The documentprovides descriptions of the methods utilized to generate the parasitic data and deriveappropriate models for their use. Some data examples, ranging from simple tabulated RLCto s-parameter models, are given to illustrate the range of electrical data that are availablefor the packages.Terminology - Definitions and ReviewsThere are a number of key concepts that should be understood in order to appreciate howpackages affect the signals transiting through them, as well as how package parasitics aremodeled or measured in the lab.Any conductor system is characterized by some basic electrical parameters which aredependent of the physical design of the system, a package is no exception. The basicelectrical parameters associated with packages are resistance, inductance, conductance,and capacitance. These are commonly referred to as RLGC parameters. The parameterswill be defined in the following subsections. The section also explains several other metricsDevice Package User GuideUG112 (v3.7) September 5, 2012Electrical Data Generation and Measurement MethodsFigure 4-5 illustrates Excel formatted tabulation of per pin data. The top 20 balls of the file for XC4VLX60 in FF1148 are shown.Models at Xilinx - Electrical Data Delivery via ModelsPackage models are a means to convey package electrical data, as stated in the previous section. These are provided to allow device users to accurately predict the performance of their designs. Xilinx recognizes that there might be several I/O model types available. For this reason, package electrical data is provided through the following I/O model formats as default:•Base [Package] section data in IBIS device file. The base [Package] section data isprovided for all newer devices in the base IBIS file. This data usually lists thepackages used with Typical, Min, and Max parasitics, as illustrated in Figure 4-4.•RLC matrix .pkg data format in either coupled RLC or uncoupled - depends on the device and size of package for SelectIO. These whole package RLC matrix datamodels are recommended for use below at about 1 Gbit/second data rates. The R, L and C element values are not frequency dependent. (The R value is typicallycharacterized at DC). The IBIS .pkg format data is intended to be read by an IBISsimulator which will utilize the data to create an appropriate package model that will be connected to the IBIS buffer model being simulated. These models can be utilized in simulators such as HyperLynx, ICX, Hspice, and others. These models are extracted Figure 4-5:Excel Formatted Tabulation of Per Pin DataUG112_C4_05_111208。
菲涅耳非相干关联全息图(综述)
Fresnel incoherent correlation hologram-a reviewInvited PaperJoseph Rosen,Barak Katz1,and Gary Brooker2∗∗1Department of Electrical and Computer Engineering,Ben-Gurion University of the Negev,P.O.Box653,Beer-Sheva84105,Israel2Johns Hopkins University Microscopy Center,Montgomery County Campus,Advanced Technology Laboratory, Whiting School of Engineering,9605Medical Center Drive Suite240,Rockville,MD20850,USA∗E-mail:rosen@ee.bgu.ac.il;∗∗e-mail:gbrooker@Received July17,2009Holographic imaging offers a reliable and fast method to capture the complete three-dimensional(3D) information of the scene from a single perspective.We review our recently proposed single-channel optical system for generating digital Fresnel holograms of3D real-existing objects illuminated by incoherent light.In this motionless holographic technique,light is reflected,or emitted from a3D object,propagates througha spatial light modulator(SLM),and is recorded by a digital camera.The SLM is used as a beam-splitter of the single-channel incoherent interferometer,such that each spherical beam originated from each object point is split into two spherical beams with two different curve radii.Incoherent sum of the entire interferences between all the couples of spherical beams creates the Fresnel hologram of the observed3D object.When this hologram is reconstructed in the computer,the3D properties of the object are revealed.OCIS codes:100.6640,210.4770,180.1790.doi:10.3788/COL20090712.0000.1.IntroductionHolography is an attractive imaging technique as it offers the ability to view a complete three-dimensional (3D)volume from one image.However,holography is not widely applied to the regime of white-light imaging, because white-light is incoherent and creating holograms requires a coherent interferometer system.In this review, we describe our recently invented method of acquiring incoherent digital holograms.The term incoherent digi-tal hologram means that incoherent light beams reflected or emitted from real-existing objects interfere with each other.The resulting interferogram is recorded by a dig-ital camera and digitally processed to yield a hologram. This hologram is reconstructed in the computer so that 3D images appear on the computer screen.The oldest methods of recording incoherent holograms have made use of the property that every incoherent ob-ject is composed of many source points,each of which is self-spatial coherent and can create an interference pattern with light coming from the point’s mirrored image.Under this general principle,there are vari-ous types of holograms[1−8],including Fourier[2,6]and Fresnel holograms[3,4,8].The process of beam interfering demands high levels of light intensity,extreme stability of the optical setup,and a relatively narrow bandwidth light source.More recently,three groups of researchers have proposed computing holograms of3D incoherently illuminated objects from a set of images taken from differ-ent points of view[9−12].This method,although it shows promising prospects,is relatively slow since it is based on capturing tens of scene images from different view angles. Another method is called scanning holography[13−15],in which a pattern of Fresnel zone plates(FZPs)scans the object such that at each and every scanning position, the light intensity is integrated by a point detector.The overall process yields a Fresnel hologram obtained as a correlation between the object and FZP patterns.How-ever,the scanning process is relatively slow and is done by mechanical movements.A similar correlation is ac-tually also discussed in this review,however,unlike the case of scanning holography,our proposed system carries out a correlation without movement.2.General properties of Fresnel hologramsThis review concentrates on the technique of incoher-ent digital holography based on single-channel incoher-ent interferometers,which we have been involved in their development recently[16−19].The type of hologram dis-cussed here is the digital Fresnel hologram,which means that a hologram of a single point has the form of the well-known FZP.The axial location of the object point is encoded by the Fresnel number of the FZP,which is the technical term for the number of the FZP rings along the given radius.To understand the operation principle of any general Fresnel hologram,let us look on the difference between regular imaging and holographic systems.In classical imaging,image formation of objects at different distances from the lens results in a sharp image at the image plane for objects at only one position from the lens,as shown in Fig.1(a).The other objects at different distances from the lens are out of focus.A Fresnel holographic system,on the other hand,as depicted in Fig.1(b),1671-7694/2009/120xxx-08c 2009Chinese Optics Lettersprojects a set of rings known as the FZP onto the plane of the image for each and every point at every plane of the object being viewed.The depth of the points is en-coded by the density of the rings such that points which are closer to the system project less dense rings than distant points.Because of this encoding method,the 3D information in the volume being imaged is recorded into the recording medium.Thus once the patterns are decoded,each plane in the image space reconstructed from a Fresnel hologram is in focus at a different axial distance.The encoding is accomplished by the presence of a holographic system in the image path.At this point it should be noted that this graphical description of pro-jecting FZPs by every object point actually expresses the mathematical two-dimensional (2D)correlation (or convolution)between the object function and the FZP.In other words,the methods of creating Fresnel holo-grams are different from each other by the way they spatially correlate the FZP with the scene.Another is-sue to note is that the correlation should be done with a FZP that is somehow “sensitive”to the axial locations of the object points.Otherwise,these locations are not encoded into the hologram.The system described in this review satisfies the condition that the FZP is depen-dent on the axial distance of each and every objectpoint.parison between the Fresnel holography principle and conventional imaging.(a)Conventional imaging system;(b)fresnel holographysystem.Fig.2.Schematic of FINCH recorder [16].BS:beam splitter;L is a spherical lens with focal length f =25cm;∆λindicates a chromatic filter with a bandwidth of ∆λ=60nm.This means that indeed points,which are closer to the system,project FZP with less cycles per radial length than distant points,and by this condition the holograms can actually image the 3D scene properly.The FZP is a sum of at least three main functions,i.e.,a constant bias,a quadratic phase function and its complex conjugate.The object function is actually corre-lated with all these three functions.However,the useful information,with which the holographic imaging is real-ized,is the correlation with just one of the two quadratic phase functions.The correlation with the other quadratic phase function induces the well-known twin image.This means that the detected signal in the holographic system contains three superposed correlation functions,whereas only one of them is the required correlation between the object and the quadratic phase function.Therefore,the digital processing of the detected image should contain the ability to eliminate the two unnecessary terms.To summarize,the definition of Fresnel hologram is any hologram that contains at least a correlation (or convolu-tion)between an object function and a quadratic phase function.Moreover,the quadratic phase function must be parameterized according to the axial distance of the object points from the detection plane.In other words,the number of cycles per radial distance of each quadratic phase function in the correlation is dependent on the z distance of each object point.In the case that the object is illuminated by a coherent wave,this correlation is the complex amplitude of the electromagnetic field directly obtained,under the paraxial approximation [20],by a free propagation from the object to the detection plane.How-ever,we deal here with incoherent illumination,for which an alternative method to the free propagation should be applied.In fact,in this review we describe such method to get the desired correlation with the quadratic phase function,and this method indeed operates under inco-herent illumination.The discussed incoherent digital hologram is dubbed Fresnel incoherent correlation hologram (FINCH)[16−18].The FINCH is actually based on a single-channel on-axis incoherent interferometer.Like any Fresnel holography,in the FINCH the object is correlated with a FZP,but the correlation is carried out without any movement and without multiplexing the image of the scene.Section 3reviews the latest developments of the FINCH in the field of color holography,microscopy,and imaging with a synthetic aperture.3.Fresnel incoherent correlation holographyIn this section we describe the FINCH –a method of recording digital Fresnel holograms under incoher-ent illumination.Various aspects of the FINCH have been described in Refs.[16-19],including FINCH of re-flected white light [16],FINCH of fluorescence objects [17],a FINCH-based holographic fluorescence microscope [18],and a hologram recorder in a mode of a synthetic aperture [19].We briefly review these works in the current section.Generally,in the FINCH system the reflected incoher-ent light from a 3D object propagates through a spatial light modulator (SLM)and is recorded by a digital cam-era.One of the FINCH systems [16]is shown in Fig.2.White-light source illuminates a 3D scene,and the reflected light from the objects is captured by a charge-coupled device (CCD)camera after passing through a lens L and the SLM.In general,we regard the system as an incoherent interferometer,where the grating displayed on the SLM is considered as a beam splitter.As is com-mon in such cases,we analyze the system by following its response to an input object of a single infinitesimal point.Knowing the system’s point spread function (PSF)en-ables one to realize the system operation for any general object.Analysis of a beam originated from a narrow-band infinitesimal point source is done by using Fresnel diffraction theory [20],since such a source is coherent by definition.A Fresnel hologram of a point object is obtained when the two interfering beams are two spherical beams with different curvatures.Such a goal is achieved if the SLM’s reflection function is a sum of,for instance,constant and quadratic phase functions.When a plane wave hits the SLM,the constant term represents the reflected plane wave,and the quadratic phase term is responsible for the reflected spherical wave.A point source located at some distance from a spher-ical positive lens induces on the lens plane a diverging spherical wave.This wave is split by the SLM into two different spherical waves which propagate toward the CCD at some distance from the SLM.Consequently,in the CCD plane,the intensity of the recorded hologram is a sum of three terms:two complex-conjugated quadratic phase functions and a constant term.This result is the PSF of the holographic recording system.For a general 3D object illuminated by a narrowband incoherent illumination,the intensity of the recorded hologram is an integral of the entire PSFs,over all object intensity points.Besides a constant term,thehologramFig.3.(a)Phase distribution of the reflection masks dis-played on the SLM,with θ=0◦,(b)θ=120◦,(c)θ=240◦.(d)Enlarged portion of (a)indicating that half (randomly chosen)of the SLM’s pixels modulate light with a constant phase.(e)Magnitude and (f)phase of the final on-axis digi-tal hologram.(g)Reconstruction of the hologram of the three characters at the best focus distance of ‘O’.(h)Same recon-struction at the best focus distance of ‘S’,and (i)of ‘A’[16].expression contains two terms of correlation between an object and a quadratic phase,z -dependent,function.In order to remain with a single correlation term out of the three terms,we follow the usual procedure of on-axis digital holography [14,16−19].Three holograms of the same object are recorded with different phase con-stants.The final hologram is a superposition of the three holograms containing only the desired correlation between the object function and a single z -dependent quadratic phase.A 3D image of the object can be re-constructed from the hologram by calculating theFresnelFig.4.Schematics of the FINCH color recorder [17].L 1,L 2,L 3are spherical lenses and F 1,F 2are chromaticfilters.Fig.5.(a)Magnitude and (b)phase of the complex Fres-nel hologram of the dice.Digital reconstruction of the non-fluorescence hologram:(c)at the face of the red dots on the die,and (d)at the face of the green dots on the die.(e)Magnitude and (f)phase of the complex Fresnel hologram of the red dots.Digital reconstruction of the red fluorescence hologram:(g)at the face of the red dots on the die,and (h)at the face of the green dots on the die.(i)Magnitude and (j)phase of the complex Fresnel hologram of the green dots.Digital reconstruction of the green fluorescence hologram:(k)at the face of the red dots on the die,and (l)at the face of the green dots on the position of (c),(g),(k)and that of (d),(h),(l)are depicted in (m)and (n),respectively [17].Fig.6.FINCHSCOPE schematic in uprightfluorescence microscope[18].propagation formula.The system shown in Fig.2has been used to record the three holograms[16].The SLM has been phase-only, and as so,the desired sum of two phase functions(which is no longer a pure phase)cannot be directly displayed on this SLM.To overcome this obstacle,the quadratic phase function has been displayed randomly on only half of the SLM pixels,and the constant phase has been displayed on the other half.The randomness in distributing the two phase functions has been required because organized non-random structure produces unnecessary diffraction orders,therefore,results in lower interference efficiency. The pixels are divided equally,half to each diffractive element,to create two wavefronts with equal energy.By this method,the SLM function becomes a good approx-imation to the sum of two phase functions.The phase distributions of the three reflection masks displayed on the SLM,with phase constants of0◦,120◦and240◦,are shown in Figs.3(a),(b)and(c),respectively.Three white-on-black characters i th the same size of 2×2(mm)were located at the vicinity of rear focal point of the lens.‘O’was at z=–24mm,‘S’was at z=–48 mm,and‘A’was at z=–72mm.These characters were illuminated by a mercury arc lamp.The three holo-grams,each for a different phase constant of the SLM, were recorded by a CCD camera and processed by a computer.Thefinal hologram was calculated accord-ing to the superposition formula[14]and its magnitude and phase distributions are depicted in Figs.3(e)and (f),respectively.The hologram was reconstructed in the computer by calculating the Fresnel propagation toward various z propagation distances.Three different recon-struction planes are shown in Figs.3(g),(h),and(i).In each plane,a different character is in focus as is indeed expected from a holographic reconstruction of an object with a volume.In Ref.[17],the FINCH has been capable to record multicolor digital holograms from objects emittingfluo-rescent light.Thefluorescent light,specific to the emis-sion wavelength of variousfluorescent dyes after excita-tion of3D objects,was recorded on a digital monochrome camera after reflection from the SLM.For each wave-length offluorescent emission,the camera sequentially records three holograms reflected from the SLM,each with a different phase factor of the SLM’s function.The three holograms are again superposed in the computer to create a complex-valued Fresnel hologram of eachflu-orescent emission without the twin image problem.The holograms for eachfluorescent color are further combined in a computer to produce a multicoloredfluorescence hologram and3D color image.An experiment showing the recording of a colorfluo-rescence hologram was carried out[17]on the system in Fig. 4.The phase constants of0◦,120◦,and240◦were introduced into the three quadratic phase functions.The magnitude and phase of thefinal complex hologram,su-perposed from thefirst three holograms,are shown in Figs.5(a)and(b),respectively.The reconstruction from thefinal hologram was calculated by using the Fresnel propagation formula[20].The results are shown at the plane of the front face of the front die(Fig.5(c))and the plane of the front face of the rear die(Fig.5(d)).Note that in each plane a different die face is in focus as is indeed expected from a holographic reconstruction of an object with a volume.The second three holograms were recorded via a redfilter in the emissionfilter slider F2 which passed614–640nmfluorescent light wavelengths with a peak wavelength of626nm and a full-width at half-maximum,of11nm(FWHM).The magnitude and phase of thefinal complex hologram,superposed from the‘red’set,are shown in Figs.5(e)and(f),respectively. The reconstruction results from thisfinal hologram are shown in Figs.5(g)and(h)at the same planes as those in Figs.5(c)and(d),respectively.Finally,an additional set of three holograms was recorded with a greenfilter in emissionfilter slider F2,which passed500–532nmfluo-rescent light wavelengths with a peak wavelength of516 nm and a FWHM of16nm.The magnitude and phase of thefinal complex hologram,superposed from the‘green’set,are shown in Figs.5(i)and(j),respectively.The reconstruction results from thisfinal hologram are shown in Figs.5(k)and(l)at the same planes as those in Fig. 5(c)and(d),positions of Figs.5(c), (g),and(k)and Figs.5(d),(h),and(l)are depicted in Figs.5(m)and(n),respectively.Note that all the colors in Fig.5(colorful online)are pseudo-colors.These last results yield a complete color3D holographic image of the object including the red and greenfluorescence. While the optical arrangement in this demonstration has not been optimized for maximum resolution,it is im-portant to recognize that even with this simple optical arrangement,the resolution is good enough to image the fluorescent emissions with goodfidelity and to obtain good reflected light images of the dice.Furthermore, in the reflected light images in Figs.5(c)and(m),the system has been able to detect a specular reflection of the illumination from the edge of the front dice. Another system to be reviewed here is thefirst demon-stration of a motionless microscopy system(FINCH-SCOPE)based upon the FINCH and its use in record-ing high-resolution3Dfluorescent images of biological specimens[18].By using high numerical aperture(NA) lenses,a SLM,a CCD camera,and some simplefilters, FINCHSCOPE enables the acquisition of3D microscopic images without the need for scanning.A schematic diagram of the FINCHSCOPE for an upright microscope equipped with an arc lamp sourceFig.7.FINCHSCOPE holography of polychromatic beads.(a)Magnitude of the complex hologram 6-µm beads.Images reconstructed from the hologram at z distances of (b)34µm,(c)36µm,and (d)84µm.Line intensity profiles between the beads are shown at the bottom of panels (b)–(d).(e)Line intensity profiles along the z axis for the lower bead from reconstructed sections of a single hologram (line 1)and from a widefield stack of the same bead (28sections,line 2).Beads (6µm)excited at 640,555,and 488nm with holograms reconstructed (f)–(h)at plane (b)and (j)–(l)at plane (d).(i)and (m)are the combined RGB images for planes (b)and (d),respectively.(n)–(r)Beads (0.5µm)imaged with a 1.4-NA oil immersion objective:(n)holographic camera image;(o)magnitude of the complex hologram;(p)–(r)reconstructed image at planes 6,15,and 20µm.Scale bars indicate image size [18].Fig.8.FINCHSCOPE fluorescence sections of pollen grains and Convallaria rhizom .The arrows point to the structures in the images that are in focus at various image planes.(b)–(e)Sections reconstructed from a hologram of mixed pollen grains.(g)–(j)Sections reconstructed from a hologram of Convallaria rhizom .(a),(f)Magnitudes of the complex holograms from which the respective image planes are reconstructed.Scale bars indicate image size [18].is shown in Fig. 6.The beam of light that emerges from an infinity-corrected microscope objective trans-forms each point of the object being viewed into a plane wave,thus satisfying the first requirement of FINCH [16].A SLM and a digital camera replace the tube lens,reflec-tive mirror,and other transfer optics normally present in microscopes.Because no tube lens is required,infinity-corrected objectives from any manufacturer can be used.A filter wheel was used to select excitation wavelengths from a mercury arc lamp,and the dichroic mirror holder and the emission filter in the microscope were used to direct light to and from the specimen through an infinity-corrected objective.The ability of the FINCHSCOPE to resolve multicolor fluorescent samples was evaluated by first imaging poly-chromatic fluorescent beads.A fluorescence bead slidewith the beads separated on two separate planes was con-structed.FocalCheck polychromatic beads(6µm)were used to coat one side of a glass microscope slide and a glass coverslip.These two surfaces were juxtaposed and held together at a distance from one another of∼50µm with optical cement.The beads were sequentially excited at488-,555-,and640-nm center wavelengths(10–30nm bandwidths)with emissions recorded at515–535,585–615,and660–720nm,respectively.Figures7(a)–(d) show reconstructed image planes from6µm beads ex-cited at640nm and imaged on the FINCHSCOPE with a Zeiss PlanApo20×,0.75NA objective.Figure7(a) shows the magnitude of the complex hologram,which contains all the information about the location and in-tensity of each bead at every plane in thefield.The Fresnel reconstruction from this hologram was selected to yield49planes of the image,2-µm apart.Two beads are shown in Fig.7(b)with only the lower bead exactly in focus.Figure7(c)is2µm into thefield in the z-direction,and the upper bead is now in focus,with the lower bead slightly out of focus.The focal difference is confirmed by the line profile drawn between the beads, showing an inversion of intensity for these two beads be-tween the planes.There is another bead between these two beads,but it does not appear in Figs.7(b)or(c) (or in the intensity profile),because it is48µm from the upper bead;it instead appears in Fig.7(d)(and in the line profile),which is24sections away from the section in Fig.7(c).Notice that the beads in Figs.7(b)and(c)are no longer visible in Fig.7(d).In the complex hologram in Fig.7(a),the small circles encode the close beads and the larger circles encode the distant central bead. Figure7(e)shows that the z-resolution of the lower bead in Fig.7(b),reconstructed from sections created from a single hologram(curve1),is at least comparable to data from a widefield stack of28sections(obtained by moving the microscope objective in the z-direction)of the same field(curve2).The co-localization of thefluorescence emission was confirmed at all excitation wavelengths and at extreme z limits,as shown in Figs.7(f)–(m)for the 6-µm beads at the planes shown in Figs.7(b)((f)–(i)) and(d)((j)–(m)).In Figs.7(n)–(r),0.5-µm beads imaged with a Zeiss PlanApo×631.4NA oil-immersion objective are shown.Figure7(n)presents one of the holo-grams captured by the camera and Fig.7(o)shows the magnitude of the complex hologram.Figures7(p)–(r) show different planes(6,15,and20µm,respectively)in the bead specimen after reconstruction from the complex hologram of image slices in0.5-µm steps.Arrows show the different beads visualized in different z image planes. The computer reconstruction along the z-axis of a group offluorescently labeled pollen grains is shown in Figs. 8(b)–(e).As is expected from a holographic reconstruc-tion of a3D object with volume,any number of planes can be reconstructed.In this example,a different pollen grain was in focus in each transverse plane reconstructed from the complex hologram whose magnitude is shown in Fig.8(a).In Figs.8(b)–(e),the values of z are8,13, 20,and24µm,respectively.A similar experiment was performed with the autofluorescent Convallaria rhizom and the results are shown in Figs.8(g)–(j)at planes6, 8,11,and12µm.The most recent development in FINCH is a new lens-less incoherent holographic system operating in a syn-thetic aperture mode[19].Synthetic aperture is a well-known super-resolution technique which extends the res-olution capabilities of an imaging system beyond thetheoretical Rayleigh limit dictated by the system’s ac-tual ing this technique,several patternsacquired by an aperture-limited system,from variouslocations,are tiled together to one large pattern whichcould be captured only by a virtual system equippedwith a much wider synthetic aperture.The use of optical holography for synthetic apertureis usually restricted to coherent imaging[21−23].There-fore,the use of this technique is limited only to thoseapplications in which the observed targets can be illu-minated by a laser.Synthetic aperture carried out by acombination of several off-axis incoherent holograms inscanning holographic microscopy has been demonstratedby Indebetouw et al[24].However,this method is limitedto microscopy only,and although it is a technique ofrecording incoherent holograms,a specimen should alsobe illuminated by an interference pattern between twolaser beams.Our new scheme of holographic imaging of incoher-ently illuminated objects is dubbing a synthetic aperturewith Fresnel elements(SAFE).This holographic lens-less system contains only a SLM and a digital camera.SAFE has an extended synthetic aperture in order toimprove the transverse and axial resolutions beyond theclassic limitations.The term synthetic aperture,in thepresent context,means time(or space)multiplexing ofseveral Fresnel holographic elements captured from vari-ous viewpoints by a system with a limited real aperture.The synthetic aperture is implemented by shifting theSLM-camera set,located across thefield of view,be-tween several viewpoints.At each viewpoint,a differentmask is displayed on the SLM,and a single element ofthe Fresnel hologram is recorded(Fig.9).The variouselements,each of which is recorded by the real aperturesystem during the capturing time,are tiled together sothat thefinal mosaic hologram is effectively consideredas being captured from a single synthetic aperture,whichis much wider than the actual aperture.An example of such a system with the synthetic aper-ture three times wider than the actual aperture can beseen in Fig.9.For simplicity of the demonstration,the synthetic aperture was implemented only along thehorizontal axis.In principle,this concept can be gen-eralized for both axes and for any ratio of synthetic toactual apertures.Imaging with the synthetic apertureis necessary for the cases where the angular spectrumof the light emitted from the observed object is widerthan the NA of a given imaging system.In the SAFEshown in Fig.9,the SLM and the digital camera movein front of the object.The complete Fresnel hologramof the object,located at some distance from the SLM,isa mosaic of three holographic elements,each of which isrecorded from a different position by the system with thereal aperture of the size A x×A y.The complete hologram tiled from the three holographic Fresnel elements has thesynthetic aperture of the size3(·A x×A y)which is three times larger than the real aperture at the horizontal axis.The method to eliminate the twin image and the biasterm is the same as that has been used before[14,16−18];。
铸造-冷挤压成形铜包铝线工艺及微观组织
第17卷第4期2010年8月塑性工程学报JOURNAL OF PLAST ICITY ENGINEERINGVol 17 No 4Aug 2010doi:10 3969/j issn 1007 2012 2010 04 005铸造 冷挤压成形铜包铝线工艺及微观组织(燕山大学,亚稳材料制备技术与科学国家重点实验室,机械工程学院,秦皇岛 066004)张春祥1 骆俊廷3(河北农业大学海洋学院,秦皇岛 066004) 王守玉2摘 要:采用铸造 冷挤压工艺制备铜包铝复合线材。
通过对制品的微观组织进行分析表明,在锥形变形区,界面处铝晶粒均开始得到细化,而中心晶粒直径较大,基本保持原态,组织极不均匀,区域分界非常明显;铜铝边界上铜的晶粒也明显得到细化,铜的中心晶粒直径较大,但两区域形态均为等轴形晶粒,变形体外表面铜的微观组织最为细小,但晶粒形状不很规则。
制品全挤出后,铜铝晶粒基本得到完全细化,而且非常细小均匀,铝芯平均直径小于5 m,铜皮平均晶粒直径在10 m 左右。
界面相晶粒尺寸非常细小均匀,与铜铝两相结合良好。
关键词:铜包铝线;铸造 冷挤压;微观组织中图分类号:T G376 3 文献标识码:A 文章编号:1007 2012(2010)04 0023 04Fabrication technology and microstructure of aluminum/copperclad composites formed by casting cold extrusionZH AN G Chun x iang 1 L U O Jun ting 3(State K ey L abo rato ry of M etastable M ater ials Science and T echno log y,Colleg e of M echanical Eng ineer ing,Y anshan U niversity,Q inhuangdao 066004 China)WA N G Shou yu 2(O cean Colleg e of H ebei Ag ricultur al U niv ersity ,Q inhuangdao 066004 China)Abstract:A luminum/co pper clad co mpo site was fabricated by casting cold ext rusion for ming techno lo gy and micro structure of pr oducts w as obser ved and analy zed.T he results show that aluminum g rains at the inter face are refined in co ne shaped defor mation zo ne,but the gr ains in the center maintain the or ig inal state and g rain size is non unifor m.A clear boundary is pr esented betw een the r efined ar ea and center ar ea.In contrast,co pper g ra ins in the radial pr ofiles has been signif icantly r efined.I n the center area of the co pper,the g r ains size is bigg er than that at bo undary.On the surface of the deformable body the g rain size is the sma llest,but with irr egular gr ain mor pho log y.A fter the product is entirely ex truded,all the copper and aluminum g rains are r ef ined w ith small and unifor m mo rpholog y.In the center ar ea,the av erag e diameter o f aluminum g rains is less than 5 m,and the co pper g rain on t he sur face is about 10 m.At the inter face the g rain size is ver y small,w ith a go od combination of co pper and a luminum.Key words:aluminum/copper clad composite;casting co ld ex tr usio n;microstr ucture骆俊廷 E mail:ljt ly k@y ahoo com cn作者简介:骆俊廷,男,1976年生,燕山大学机械工程学院,副教授,博士,主要研究方向为精密塑性成形工艺、纳米材料的塑性与塑性加工技术收稿日期:2010 04 08;修订日期:2010 04 12引 言铜包铝线是以铝为芯体,在其上包覆一层铜,实现铜、铝界面固相结合的一种复合材料[1 2]。
实对称矩阵正定、半正定的简易判别
目 录1.引言 ......................................................................... 错误!未定义书签。
2.实对称矩阵正定、半正定的简易判别方法 ............. 错误!未定义书签。
2.1 实对称 矩阵的几个定义[]3 .............................................. 错误!未定义书签。
2.2 实对称矩阵正定的充分必要条件有下列几种方法: ............................................ 1 2.3 实对称矩阵正定简易判别的几个充分必要条件。
.............................................. 3 2.3.1 n 阶实对称矩阵A 正定的充分必要条件是合同A 于单位矩阵E []3. (4)2.3.2n 元实二次型正定的充分必要条件是它的正惯性指数等于[]9n 。
.................... 5 2.4 实对称矩阵半A 正定的几个充分必要条件[]6。
................................................ 5 2.4.1 二次型()n x x x f ,,,21 Ax x T=,其中A A T =,()n x x x f ,,,21 半正定。
. 52.4.2n 阶实对称矩阵A 是半正定矩阵的充分必要条件是的正惯性A 指数等于它的秩。
(5)2.4.3n 阶对称矩阵是A 半正定矩阵的充分必要条件是的特征值全A 大于等于零,但至少有一个特征值等于零。
(5)2.4.4 实对称矩阵的A 所有主子式皆大于或等于零。
............................................. 5 2.4.5 有实矩阵C 使C C A T=,则A 半正定。
雅思刘洪波538阅读词汇
accelerate
accumulate
ack no wledge
Ad aptto
addictive
adjust
admit
Adversity
aggressive
agreeable
aid
allergic
ocate
alter
Alter native
altitude
an alyze
an cie nt
comprehe nd
con ceal
concen trate
concur
condemn
confer
con fide ntial
con flict
conform
con fuse
con scious
con servative
con siderable
con stra in
con sult
con ti ngent
react
Realistic
realize
recog nize
record
Recreati on
reductio n
Referto
regulati on
Rehearsal
reject
relative
releva nt
Relia nee
religious
relucta nt
Rely on
Replace
Tran smit
Treme ndous
Trigger
Tropical
Un biased
Un derestimate
Un derlie
Un expected
Uniform
Pairing fluctuation effects on the single-particle spectra for the superconducting state
a r X i v :c o n d -m a t /0311424v 1 [c o n d -m a t .s u p r -c o n ] 18 N o v 2003Pairing fluctuation effects on the single-particle spectra for the superconducting stateP.Pieri,L.Pisani,and G.C.StrinatiDipartimento di Fisica,UdR INFM,Universit`a di Camerino,I-62032Camerino,Italy(February 2,2008)Single-particle spectra are calculated in the superconducting state for a fermionic system with an attractive interaction,as functions of temperature and coupling strength from weak to strong.The fermionic system is described by a single-particle self-energy that includes pairing-fluctuation effects in the superconducting state.The theory reduces to the ordinary BCS approximation in weak coupling and to the Bogoliubov approximation for the composite bosons in strong coupling.Several features of the single-particle spectral function are shown to compare favorably with experimental data for cuprate superconductors.PACS numbers:03.75.Ss,03.75.Hh,05.30.JpInformation on the single-particle spectral function,that is obtained from ARPES 1and tunneling data 2for cuprate superconductors,can shed light on the charac-teristic features of the superconducting state as well as on its connection with the unconventional normal state above the critical temperature T c .Most prominent among these features are the continu-ous evolution of a broad pseudogap structure from above to below T c ,1,2the emergence of a coherent peak below T c that combines with the pseudogap structure to yield a characteristic peak-dip-hump profile 1(for which two distinct energy scales can be identified),and the pecu-liar dependence of the frequency position and weight of the coherent peak on temperature and doping.1Gener-ally speaking,features of the standard BCS theory are recovered for overdoped samples,while non-BCS behav-iors occur for optimally-doped and underdoped samples.The origin of the peak-dip-hump profile has especially been the subject of controversy,being attributed either to “extrinsic”effects like the bilayer splitting 3or to “intrin-sic”effects.The latter can be identified over and above the extrinsic effects,and are believed to originate from strong (many-body)interactions in the system 4.Quite independently from the microscopic origin of the fermionic attraction giving rise to superconductiv-ity,its strength is believed to be stronger in optimally-doped and underdoped than in overdoped samples 5,con-sistently with the above findings.This implies that pair-ing fluctuations should definitely be taken into account,irrespective of other effects (such as the bilayer splitting and/or additional many-body effects associated with spe-cific pairing mechanisms 6).In this paper,we assess the role of pairing fluctua-tions for the single-particle spectral function in the su-perconducting state on rather general grounds,by iden-tifying a single-particle self-energy that describes fluctu-ating Cooper pairs in weak coupling and non-condensed composite bosons in strong coupling.(The latter form as bound-fermion pairs in the strong-coupling limit of the fermionic attraction.)To this end,we consider fermions mutually interacting via an attractive contact potential in a 3D continuum,without taking into account lattice effects nor the explicit physical mechanism responsible for the attraction.Qualitative comparison with experi-mental data will thus rest on identifying the increasing coupling strength in this model with the increasing po-tential strength for decreasing doping level in the phase diagram of cuprate superconductors.Although the ef-fective model here considered is oversimplified for a full description of cuprates,the physical questions we are ad-dressing are sufficiently general that this minimal model will prove sufficient to capture the main experimental features.Our main results for the single-particle spectral func-tion (to be compared with the experimental findings)are summarized as follows:(i)A broad pseudogap structure and a coherent peak are simultaneously present in a wide range of coupling and temperature below T c ,giving rise to a peak-dip-hump profile.Two distinct energy scales (the positions ∆pg of the broad pseudogap structure and ∆m of the coherent peak)can then be extracted from the spectra at given temperature and coupling,as seen experimentally 1.(ii)At a given (notably,intermediate)coupling,BCS-like features coexist with non-BCS behaviors.The position ∆m and weight z of the coherent peak versus wave vec-tor follow a BCS-like dependence,as found experimen-tally 7,8.At the same time,the weight z of the coherent peak has a strong temperature dependence,in agreement with experiments 9,10but in contrast with BCS theory.(iii)At low temperature,the weight z of the coherent peak has a strong dependence on coupling,decreasing monotonically from weak to strong coupling (as seen ex-perimentally in cuprates for decreasing doping 9,10).At the same time,the position ∆m of the coherent peak in-creases monotonically with coupling 9,10.(iv)The positions ∆pg of the broad pseudogap structure at T c and ∆m of the coherent peak near zero temperaturecross each other as function of coupling about interme-diate coupling.This feature is also seen experimentally by intrinsic tunneling experiments at different dopings11 (although these data are subject to controversy12). Pairingfluctuations are taken into account in our the-ory by considering,besides the off-diagonal BCS-like self-energy,the diagonal t-matrix self-energy suitably ex-tended to the superconducting state(we use the Nambu formalism throughout).The t-matrix self-energy has been widely used to include pairingfluctuations in the normal state13,and specifically to account for pseudogap features in the single-particle spectral function14.In our theory,the diagonal t-matrix self-energy(that survives above T c)will essentially be responsible for the pres-ence of the broad pseudogap structure below T c.The off-diagonal BCS-like self-energy will instead give rise to the simultaneous emergence of the coherent peak.In ad-dition,our theory recovers the Bogoliubov approximation for the composite bosons in strong coupling.15The diagonal t-matrix self-energy readsΣ11(k)=− d qβ ΩνΓ11(q)G11(q−k),(1)while the off-diagonal self-energy has the BCS-like form Σ12(k)=−∆in terms of the superconducting order pa-rameter∆.The pairing-fluctuation propagatorΓ11in the broken-symmetry phase entering Eq.(1)is given by:Γ11(q)=χ11(−q)4πa F+ d pβ ωn G11(p+q)G11(−p)−m(2π)31(2π)314πa F = d k2E(k)−miωl+ξ(k)+Σ11(k,−ωl) −1.(7)Note that,while the number equation(5)containsthe dressed normal(diagonal)Green’s function,thegap equation(6)is obtained from the anomalous(off-diagonal)BCS Green’s function and thus its form is notmodified with respect to the BCS theory.This ensuresthat the pairing-fluctuation propagator(2)remains gap-less for all temperatures(below T c)and couplings.[Thevalue of T c is obtained from Eqs.(5)and(6)by setting∆=0identically.]The numerical values of the chemi-cal potentialµand order parameter∆at given tempera-ture and coupling differ,however,from those obtained byBCS theory due to the different structure of the numberequation.It can be verified15from Eqs.(2)-(6)that theBogoliubov approximation for the composite bosons is re-covered in the strong-coupling(βµ→−∞and∆≪|µ|)limit.This represents a notable achievement of our the-ory.The use of bare BCS single-particle Green’s functionsin the self-energy(1)further enables us to perform theanalytic continuation to real frequencyωin a closed form,as to avoid numerical extrapolation procedures.To thisend,and similarly to what was done in Ref.14for thenormal phase,we express the pairing-fluctuation propa-gatorΓ11by its spectral representation:Γ11(q,Ων)=−1iΩν−ω′.(8)Here ImΓ11(q,ω)is defined as the imaginary part ofΓ11(q,iΩν→ω+iη).This quantity is,in turn,ob-tained from the expressions(2)-(4),with the replacementiΩν→ω+iηmade after the sum over the internal(Mat-subara)frequency has been performed.In this way,theimaginary part of the retarded self-energyΣR11is:ImΣR 11(k ,ω)=−d qπIm G R11(k ,ω)of interest iseventually obtained as a function of ωfor any given k .160.511.522.533.54-0.4-0.200.20.4A (k µ,,ω)εFω/εFT/T c =1.00.80.70.50.1FIG.1.Single-particle spectral function vs frequency (in units of εF )at different temperatures,for |k |=k µ′and (k F a F )−1=−0.45.Figure 1shows the evolution with temperature of the single-particle spectral function for an intermediate cou-pling ((k F a F )−1=−0.45).[For this coupling,the ra-tio of the pair-breaking temperature T ∗to the critical temperature T c is about 1.25,as obtained in Ref.14.]Focusing specifically on the features at negative ω,note how the (sharp)coherent peak grows from the (broad)pseudogap structure already present at T c .The coher-ent peak becomes sharper upon lowering the tempera-ture and gains weigth at the expenses of the pseudogap structure,giving rise to a characteristic peak-dip-hump profile.The two features coexist over a wide range of temperature.Figure 2shows the wave-vector dependence of the single-particle spectral function at the temperature T =0.7T c for the same coupling of Fig.1.Following the evolu-tion of the coherent peak across the underlying Fermi sur-face,one identifies the characteristic particle-hole mix-ing of the BCS theory,with a reflection of both particle and hole bands accompanied by a transfer of the spectral weight from negative to positive frequencies.In our the-ory,such BCS-like features coexist with non-BCS behav-iors,namely,the occurrence of the pseudogap structureand the strong temperature dependence of the spectralweight of the coherent peak (see Fig.4below).123456-0.8-0.6-0.4-0.200.20.40.60.8A (k ,ω)εFω/εFk/k µ,=0.60.81.01.21.4FIG.2.Single-particle spectral function vs frequency (in units of εF )at different wave vectors about k µ′,for T =0.7T c and (k F a F )−1=−0.45.12345-2-1.5-1-0.500.511.52∆m /εF(k F a F )-1(b)0.00.10.20.30.40.5z(a)FIG.3.(a)Weight z and (b)position ∆m (full line)of the coherent peak at negative frequencies vs coupling for T =0.1T c .In (b)the pseudogap ∆pg for T =T c (dashed line)is also shown.Figures 3(a)and 3(b)report,respectively,the weight z and position ∆m of the coherent peak of the single-particle spectral function at negative frequencies vs cou-pling for a low temperature (T =0.1T c ).The peak weight z saturates at the BCS value 0.5in weak cou-pling,decreases markedly across the crossover region−1<∼(k F a F )−1<∼+1,and becomes negligible in strongcoupling.At the same time,the peak position ∆m in-creases monotonically across the crossover region.In Fig.3(b)we also report the pseudogap ∆pg (dashed line),as identified from the position of the maximum of the spectral function at T c .While the qualitative trend of ∆m and ∆pg vs coupling is similar,the two curves cross each other at about the intermediate-coupling value (k F a F )−1=−0.45.In addition,we have verified that ∆m about coincides with the value of the order parame-ter ∆in the weak-to-intermediate coupling region.00.20.40.60.810.10.30.50.70.9z (T )/z (0)T/T cFIG.4.Temperature dependence of the weight z of the coherent peak at negative frequencies (full line)for (k F a F )−1=−0.45.The superfluid density ρs (dashed line)is also shown for comparison.Figure 4reports the dependence of the weight z on temperature for the coupling value (k F a F )−1=−0.45(full line).Note the strong temperature dependence of this quantity,which vanishes at T c .This contrasts the BCS behavior,whereby the weight of the coherent peak for negative frequencies would equal 0.5irrespective of temperature.The temperature dependence of the su-perfluid density ρs (calculated according to Ref.17)is also shown in the figure (dashed line).The resemblance between these two quantities has been noted experimen-tally for nearly-optimally-doped cuprates 9,10.We have,however,verified that this resemblance does not occur for other values of the coupling,both on the weak-and strong-coupling sides of the crossover.On the weak-coupling side of the crossover,in particular,our finding is confirmed by the BCS theory,whereby z =0.5irre-spective of temperature while ρs decreases monotonically from T =0to T c .For this reason,no universal corre-spondence between the temperature dependence of z and ρs should be expected on physical grounds.The results shown in the above figures refer mostly to the weak-to-intermediate coupling side of the crossover region.Our conclusions have,however,been drawn from a sistematic study of the whole crossover region from weak to strong coupling,of which Fig.3is an example.All the qualitative features extracted from the above figures compare favorably with the experimental data oncuprates,as anticipated in the points (i)-(iv).This qual-itative comparison rests on the assumed correspondence between the increasing of the coupling strength in the present theory and the decreasing doping level in the phase diagram of cuprate superconductors.In conclusion,we have shown that pairing fluctuations can (at least qualitatively)account for several nontrivial features of single-particle spectra in the superconduct-ing state.Our results specifically demonstrate that the experimental finding of two distinct features (pseudogap structure and coherent peak)in the single-particle spec-tral function is fully consistent with the occurrence of strong pairing fluctuations in cuprate petition of two distinct order parameters is there-fore not required to account for the occurrence of two different energy scales in the experimental data.We are indebted to A.Perali for discussions.Financial support from the Italian MIUR under contract COFIN 2001Prot.2001023848is gratefully acknowledged.2mµ′for µ′>0with µ′=µ−Σ0.The spectra forA (k ,ω)are shown in the text for this special values of |k |.17N.Andrenacci,P.Pieri,and G.C.Strinati,Phys.Rev.B 68,144507(2003).。
NORMA 4000 5000 Power Analyzer 用户说明手册说明书
Since some countries or states do not allow limitation of the term of an implied warranty, or exclusion or limitation of incidental or consequential damages, the limitations and exclusions of this warranty may not apply to every buyer. If any provision of this Warranty is held invalid or unenforceable by a court or other decision-maker of competent jurisdiction, such holding will not affect the validity or enforceability of any other provision.
BEGRENZTE GEWÄHRLEISTUNG UND HAFTUNGSBESCHRÄNKUNG
Fluke gewährleistet, daß jedes Fluke-Produkt unter normalem Gebrauch und Service frei von Material- und Fertigungsdefekten ist. Die Garantiedauer beträgt 2 Jahre ab Versanddatum. Die Garantiedauer für Teile, Produktreparaturen und Service beträgt 90 Tage. Diese Garantie wird ausschließlich dem Erster
机械英语专业名词
机械专业英语词汇(很全)金属切削m etal c utting机床 machi ne too l 金属工艺学t echnol ogy of metal s 刀具 cut ter 摩擦 fr iction联结 link传动 driv e/tran smissi on 轴 shaf t 弹性 elas ticity频率特性f requen cy cha racter istic误差error响应 resp onse 定位a llocat ion 机床夹具jig动力学 dynam ic 运动学 ki nemati c 静力学 sta tic分析力学 anal yse me chanic s 拉伸 pull ing压缩hittin g 剪切 shea r 扭转 twis t 弯曲应力 be ndingstress强度 int ensity三相交流电 th ree-ph ase AC磁路 mag neticcircle s 变压器 tra nsform er异步电动机 asyn chrono us mot or 几何形状g eometr ical精度 preci sion 正弦形的 sinus oid 交流电路 AC ci rcuit机械加工余量machin ing al lowanc e 变形力 def orming force变形 def ormati on 应力 str ess 硬度 ri gidity热处理 he at tre atment退火 annea l 正火 norm alizin g脱碳 de carbur izatio n 渗碳 carb urizat ion电路circui t 半导体元件s emicon ductor eleme nt反馈f eedbac k 发生器 gen erator直流电源 DC el ectric al sou rce门电路 gatecircui t 逻辑代数 lo gic al gebra外圆磨削e xterna l grin ding内圆磨削 int ernalgrindi ng 平面磨削planegrindi ng变速箱gearbo x 离合器 clu tch 绞孔 fr aising绞刀 rea mer 螺纹加工thread proce ssing螺钉 scre w 铣削 mill铣刀 milli ng cut ter功率 power工件 workp iece 齿轮加工 gearmechin ing齿轮gear 主运动main m ovemen t主运动方向direct ion of mainmoveme nt进给方向 dire ctionof fee d 进给运动 fe ed mov ement合成进给运动result ant mo vement of fe ed合成切削运动 res ultant movem ent of cutti ng合成切削运动方向d irecti on ofresult ant mo vement of cu tting切削深度 cu ttingdepth前刀面 rakeface刀尖 noseof too l 前角 rake angle后角cle arance angle龙门刨削pla ning 主轴s pindle主轴箱head stock卡盘chuc k 加工中心ma chinin g cent er 车刀lat he too l车床l athe 钻削镗削bore车削turni ng 磨床gri nder 基准b enchma rk钳工l ocksmi th 锻forg e 压模stam ping 焊we ld 拉床bro aching machi ne拉孔b roachi ng 装配ass emblin g 铸造foun d 流体动力学f luid d ynamic s流体力学fluidmechan ics 加工ma chinin g 液压hydr aulicpressu re切线t angent机电一体化m echano tronic s mech anical-elect ricalintegr ation气压airpressu re pne umatic press ure 稳定性stabil ity介质medium液压驱动泵fl uid cl utch液压泵hydr aulicpump 阀门v alve 失效i nvalid ation强度intens ity载荷load 应力s tress安全系数saft y fact or 可靠性re liabil ity螺纹thread螺旋helix键spline销pin 滚动轴承rollin g bear ing滑动轴承slid ing be aring弹簧spring制动器arre ster b rake十字结联轴节c rosshe ad 联轴器co upling链chain皮带strap精加工fin ish ma chinin g 粗加工rou gh mac hining变速箱体g earbox casin g 腐蚀rust氧化oxida tion磨损wear耐用度durab ility随机信号rand om sig nal离散信号disc rete s ignal超声传感器ult rasoni c sens or集成电路integ rate c ircuit挡板orifi ce pla te残余应力resi dual s tress套筒sleev e 扭力tor sion 冷加工cold m achini ng 电动机elect romoto r 汽缸cyl inder过盈配合int erfere nce fi t热加工h otwork摄像头CCDcamera倒角round ing ch amfer优化设计op timaldesign工业造型设计i ndustr ial mo ulding desig n有限元f initeelemen t 滚齿hobb ing 插齿ge ar sha ping伺服电机act uating motor铣床milli ng mac hine钻床drill machi ne 镗床bor ing ma chine步进电机step per mo tor丝杠screwrod 导轨le ad rai l 组件suba ssembl y可编程序逻辑控制器P rogram mableLogicContro ller P LC电火花加工elec tric s park m achini ng电火花线切割加工e lectri cal di scharg e wire - cut ting相图phase diagr am 热处理he at tre atment固态相变s olid s tate p hase c hanges有色金属n onferr ous me tal 陶瓷cer amics合成纤维synt heticfibre电化学腐蚀e lectro chemic al cor rosion车架autom otivechassi s悬架su spensi on 转向器re direct or 变速器sp eed ch anger板料冲压sh eet me tal pa rts 孔加工s pot fa cing m achini ng车间w orksho p 工程技术人员engine er 气动夹紧p neumalock数学模型mat hemati cal mo del 画法几何descri ptivegeomet ry机械制图Mecha nicaldrawin g 投影proj ection视图vie w 剖视图pro file c hart 标准件stand ard co mponen t零件图p art dr awing装配图assem bly dr awing尺寸标注si ze mar king 技术要求techn ical r equire ments刚度rigi dity 内力i nterna l forc e 位移disp laceme nt截面sectio n 疲劳极限f atigue limit断裂frac ture塑性变形pla stic d istort ion 脆性材料brittl enessmateri al刚度准则rigid ity cr iterio n 垫圈was her 垫片sp acer直齿圆柱齿轮s traigh t toot hed sp ur gea r 斜齿圆柱齿轮helica l-spur gear直齿锥齿轮s traigh t beve l gear运动简图ki nemati c sket ch齿轮齿条pinio n andrack 蜗杆蜗轮worm and w orm ge ar虚约束passiv e cons traint曲柄crank摇杆racke r凸轮ca ms 共轭曲线c onjuga te cur ve 范成法ge nerati on met hod定义域defi nition al dom ain值域range导数\微分dif ferent ial co effici ent求导deriva tion 定积分defin ite in tegral不定积分i ndefin ite in tegral曲率curva ture 偏微分partia l diff erenti al毛坯r ough 游标卡尺slide calip er 千分尺mi cromet er cal ipers攻丝tap二阶行列式se cond o rder d etermi nant逆矩阵inve rse ma trix 线性方程组line ar equ ations概率pro babili ty 随机变量r andomvariab le排列组合permu tation and c ombina tion 气体状态方程equ ationof sta te ofgas 动能kineti c ener gy 势能pot ential energ y机械能守恒conse rvatio n of m echani cal en ergy动量momen tum 桁架tr uss 轴线ax es 余子式co factor逻辑电路l ogic c ircuit触发器flip-flop脉冲波形puls e shap e数模di gitalanalog y 液压传动机构fluid drive mecha nism机械零件mec hanica l part s 淬火冷却qu ench 淬火h ardeni ng回火t emperi ng 调质har dening and t emperi ng 磨粒ab rasive grain结合剂bo ndingagent砂轮grindi ng whe el(be)qualfi ed, up to gr ade 合格abnor mal en gine n oise d iagnos is equ ipment异响诊断仪abnor mal ha ndling异常处理a bnorma l knoc king 异响abra sion w ear; s cratch ing 磨损abr asivebelt g rindin g mach ine 砂带磨床abr asivebelt g ringdi ng 砂带磨削abs cissa横坐标AC arc w elding machi ne 交流弧焊机acc elerat ed run赶点acc elerat ion 加速度accel eratio n anal ysis 加速度分析a cceler ationdiagra m 加速度曲线accid entaldefect事故性缺陷acety lene 乙炔acety lene c ylinde r 乙炔罐a cetyle ne gen erator乙炔发生器acety lene-w elding insta llatio n 乙炔氧焊装置add lubri cating oil 加润滑油add endum齿顶高add endumcircle齿顶圆a dditio n to s tandar d 标准补充件addit ionalelemen t 附加要素addit on age nt 添加剂adhesi on 粘附a djusta ble sp eed mo tors 调速电动机a djusti ng 调整a dminis tratio n/gene ral af fairsdept 总务部adm inistr ation/genera l affa irs de pt 总务部admis istrat ion of vehic le ami ntenan ce 汽车维修管理ad misist rative stand ard 管理标准adop ting b y equa tion 等同采用ad opting by eq uivale nt 等效采用adopt ing by refer ence 参照采用ad option of in ternat ionalstanda rd and advan ced ov erseastanda rd 采用国际标准和国外先进标准a fter d rippin g (柴油喷射系)渗漏滴油agei ng 老化a geing时效aims of st andard izatio n 标准化的目的air circu it bre aker 空气断路器ai r dry自然干燥a ir jar ring m ouldin g mach ine 气动震实造型机air pi pe 气管a ir sea l 气幕ai r spra y gun空气喷枪a ir spr ing 空气弹簧airtermin al 航空集散站air tool风动工具ai r-cond itioni ng equ ipment车厢空调设备air-cushio n ejec t-rod气垫顶杆ai rlesssprayequipm ent 无空气喷涂设备alcoho l 乙醇al coholcontai ner 沾湿台alcoh ols so lvent醇类溶剂a lignin g 校正al kalidi pping脱脂alka li-ear th bab bitt 碱土巴氏合金alkali ne deg reasin g 碱法脱脂alkali ne etc h 龄咬a llowab le amo unt of unbal ance 许用不平衡量a llowab le pre ssureangle许用压力角allowa ble st ress;permis siblestress许用应力a lloy c ast ir on 合金铸铁allo y stru cturea l stee l 合金结构钢alloy toolsteel合金工具钢altern ationswitch转换开关a lumel铝镍合金a lumini ed ste el she et 渗铝钢板alumi nium b ronze铝青铜al uminiu m cast ing al loy 铸造铝合金alu minizi ng 渗铝alumin um pow der pa int 铝粉漆(银色漆)amend ed dra ft 修正草案amend ment t o stan dard 标准修改件a mino p lastic s 氨基塑料amino-filler氨基腻子a mountof unb alance不平衡量a mplitu de ofvibrat ion 振幅analog-modedevice类模器an alysis of me chanis m 机构分析analy ticaldesign解析设计a nalyti cal me thod 分析法ang le ofcontac t 包角an gularaccele ration角加速度a ngular conta ct bal l bear ing 角接触球轴承a ngular conta ct bea ring 角接触轴承an gularcontac t radi al bea ring 角接触向心轴承angul ar con tact t hrustbearin g 角接触推力轴承an gularveloci ty 角速度angula r velo city r atio 角速比ani on ele ctroph oretic coati ng 阴离子型电泳涂料a nneal退火anne aling退火anne alingbox 退火箱annea ling f urnace退火炉an nealin g oven退火炉an nex 附录annula r ball barin g 径向球轴承annul ar spr ing 环形弹簧anod e mech anical worki ng 阳极机械加工An odize阳性处理an olyte阳极电解液a nticor rosive bimet al 耐蚀双金属anti corros ive pa int 防蚀漆anti-foamin g addi tive 防泡沫添加剂a ntifoa ming a gent 消泡剂ant ifouli ng pai nt(cop per pa int) 防污漆anti-freez e addi tive 防冻添加剂a ntifri ctionalloy减摩合金an tifric tion a lloy 减摩合金ant ifrict ion be arng 减摩轴承an ti-fri ctionqualit y 减摩性a nti-kn ock co mpound抗爆剂an tirust paint防锈漆a nti-ta rnishtreatm ent 抗暗处理aper iodicspeedfluctu ation非周期性速度波动app endeddescri ptionof sta ndard标准附加说明applic ationfactor工况系数applie d forc e 作用力a pprove d by 核准appr oved b y / ch eckedby / p repare d by 核准/审核/承办aqua regia王水arc weldi ng mac hine 电弧焊机Arc himede s worm阿基米德蜗杆arch ives o f stan dardiz ation标准化档案A rctan反正切arg on arc weldi ng 氩弧焊argon weldi ng 氩焊a rm 臂部a rmatur e test er 电枢试验器art ery 干线articu latedbus 通道式公共汽车a rticul ated e quipme nt 铰接装置arti culate d trol ly bus通道式无轨电车arti ficial dryin g 人工干燥artif icialgasoli ne 人造汽油artif icialleathe r 人造胶革artifi cial r esin 人造树胶ar tifici al woo l 人造羊毛asbest os 石棉a sbesto s band石棉带as bestos board石棉板a sbesto s fibe r 石棉纤维asbest os lin en 石棉布asbest os pap er 石棉纸asbes tos st eel le af 石棉钢片asbes tos we b 石棉织物as-dep osited stren gth 焊后强度ase ptic 防腐剂asse mbling clear ance 装配间隙ass embly组装ass emblycondit ion 装配条件Asse mbly l ine 组装线Assem bly li ne 组装线assis tant m anager助理ass istant manag er 助理A ssur g roup 杆组as-w elded焊态atla s 图册、图谱atomo sphere quali ty sta ndard大气质量标准atten dant s eat 乘务员座椅aug ular o ffset角度偏差au sformi ng 形变热处理aus tenite奥氏体au thorit y 权力机构autocl ino 自动壳型机au tomati c boxsplitt er 自动分箱机auto maticcloser自动合箱机autom atic c ompens atingdevice for b oringcutter wear镗刀磨损自动补偿装置a utomat ic gas cutti ng 自动气割auto maticheadst ock ch anging N/C m achine自动更换主轴箱数控机床autom atic i nspect ion ma chinefor pi ston r ing te nsion活塞环弹性自动检验机a utomat ic ins tructi on cal ling d evice自动指令呼叫装置aut omatic linefor we ightin g conn ecting rod 连杆称重去重自动线aut omatic mould ing pr oducti o line ofr c ylinde r bloc k 缸体自动造型生产线automa tic mu lti-st age co ld fro mer 多工位自动冷镦机autom atic p lating machi ne 自动电镀装置au tomati c prod uction linefor au tomobi le fin al dri ve hou sing 汽车主减速器壳加工自动线automa tic pr oducti on lin e forconnec ting r od 连杆加工自动线a utomat ic pro ductio n line for c ransha ft dyn amic b alance曲轴动平衡自动线au tomati c prod uction linefor cy linder block气缸体加工自动线au tomati c prod uction linefor cy linder head气缸盖加工自动线aut omatic produ ctionline f or gas carbu rizati on 气体渗碳自动线a utomat ic scr ewdriv er 电动启子auto maticspraye r 自动喷漆机auto maticticket check er 自动检票机aut omatic toolchangi ng N/C machi ne 自动换刀数控机床automa tic we lding自动焊接a utomat ion 自动化auto mobile trans portat ion en terpri se 汽车运输企业au tomoti ve alc quer 汽车用喷漆a utomot ive en amel 汽车漆aut omotiv e topc oat 汽车面漆aux iliary fucti on 辅助功能avai lablemateri al 良品可使用ave rage d ays du ring m ajor r epairof veh icles汽车大修平均在修车日a verage daysin pla nt dur ing ma jor of vehic les 汽车大修平均在厂车日ave rage i nterva l mile age of major repai r of v ehicle s 汽车大修间隔里程a verage main-hoursof veh icle m ainten ance a nd rep air 汽车维修平均工时avera ge str ess 平均应力ave rage t onnage (pass engerseat)平均吨(客)位aver age ve locity平均速度aviona l 铝合金a xial c ontact beari ng 轴向接触轴承ax ial di rectio n 轴向ax ial in ternal clear ance 轴向游隙ax ial lo ad 轴向载荷axial loadfactor轴向载荷系数axia l plan e 轴向平面axialthrust load轴向分力a xial t ooth p rofile轴向齿廓a xle st and 轮轴架babb it met al 巴氏合金backcone 背锥back coneangle背锥角bac k cone dista nce 背锥距back fire回火back flow回流back lash 侧隙backl ash 齿侧间隙bac klash间隙back-to-ba ck arr angeme nt 背对背安装bad condi tion o f vehi cle 汽车不良状况b affleplate挡块bain ite 贝氏体bakel ite 胶木(电木)ba king p aint 烘漆balan ce 平衡balanc e of m echani sm 机构平衡balan ce ofrotor转子平衡b alance of ro tors 回转体平衡ba lanceof sha king f orce 惯性力平衡b alanci ng cen tering machi ne 平衡-钻中心孔机床balan cing m achine平衡机ba lancin g mass平衡质量b alanci ng qua lity 平衡品质ba lancin g spee d 平衡转速balata巴拉塔树胶balata sprin g 橡胶弹簧ball球ballbearin g 球轴承b all be aring球轴承bal l scre w 滚珠丝杆bandbrake带式制动器b and-ai d 创可贴b ar sto ck cut ting b y punc hing 冲压下料ba rcode条码barc ode 条码barcod e 条码ba rcodescanne r 条码扫描器barc ode sc anner条码扫描器b arrel(cylin dric)cam 圆柱式凸轮步进运动机构ba rrel p lating滚镀bar-typecylind er gau ge 杆式气缸量规ba se cir cle 基圆base c one 基圆锥basecylind er 基圆柱basepitch基圆齿距ba sic dy namicaxialload r ating轴向基本额定动载荷ba sic dy namicradial loadrating径向基本额定动载荷b asic i nterna tional stand ard 基础国际标准b asic r atinglife 基本额定寿命b asic s tandar d 基础标准basic stati c axia l load ratin g 轴向基本额定静载荷basicstatic radia l load tatin g 径向基本额定静载荷basicterms基本术语ba sket 蝴蝶竺batc h furn ace 分批炉batt ery hy dromet er 蓄电池液体比累计b attery welde r 蓄电池式焊机beput in stora ge 入库b eacon警示灯bea d 焊道be ad cra cking焊道裂纹b ead we lding堆焊bead ing st ress 弯曲应力bea ring 轴承bear ing al loy 轴承合金bear ing bl ock 轴承座beari ng bor e diam eter 轴承内径be aringbush 轴瓦、轴承衬b earing cap 轴承盖bear ing ca pacity承载能力bearin g capa city f actor承载量系数b earing clear ance 轴承间隙be aringcone 轴承内圈bea ring c up 轴承盖bearin g cup轴承外圈b earing heigh t 轴承高度bearin g life轴承寿命b earing metal轴承合金bearin g outs ide di ameter轴承外径b earing race轴承座圈b earing saddl e 轴承座b earing score拉瓦bea ring s hell 轴瓦bear ing sh im 轴承垫片beari ng ste el 轴承钢bearin g widt h 轴承宽度beesw ax 蜂蜡b ehindschedu le 晚点b ellevi lle sp ring 碟形弹簧be lt dri ving 带传动belt pulle y 带轮be nder 弯曲机bend ing di e 压弯模bendin g mome nt 弯矩b ending press压弯机be nding-beadin g die压弯压筋模bendin g-curl ing di e 压弯卷圆模bendi ng-fla ngingdie 压弯翻边模be rylliu m bron ze 铍青铜bevelgear 锥齿轮beve l gear s 圆锥齿轮机构bev el pul ley; b evel w heel 锥轮bezel斜视规be zel pa nel 面板bezel panel面板bib liogra phy 文献书目bila terall y harm onized stand ard 双边协调标准b iliogr aphica l refe rence书目参考目录bimeta l stri p from tin-a lumini um all oy 高锡铝合金双金属带binder粘合剂bi sector平分线bi tumino us pai nt 沥青漆black box 黑箱black washe r 黑皮垫圈blacks mith w elding锻焊bl ank 齿轮轮坯blan k 轮坯bl ank 制坯blanke r 落料机b lankin g 穿落模blanki ng die落料模bl anking-coini ng die落料压印模blank ing-sl otting die 落料切槽模bl end so lvent溶合溶剂b linste r 气泡bl ock di agram框图bloc k hamm er 落锤blow-b y 串气bl ow-bymeter曲轴箱窜气量测定仪bl ower 鼓风机blue ing 发蓝处理blus h 导色b oard 看板bodybumpin g tool车身修整工具body guida nce me chanis m 刚体导引机构bod y of a norma tive d ocumen t 一个标准文件的主体bold l ine 粗线bolt 螺栓bonegum 骨胶booth喷漆室bor ing 镗boring镗削bor ing ho le 镗孔b oron s teel 硼钢boron izing渗硼bot tom cl earanc e 顶隙bo ttom d ie(cou nter d ie) 下模bound ary di mensio n 外形尺寸box fu rnace(muffle furna ce) 马弗炉boxmoldin g mach ine 有箱造型机box ing 环焊bracke t 小磁导brake制动器bra ke ble eder 制动系空气排除器brak e depr esor 制动踏板压下器brakedrum l athe 制动鼓车床b rake f ade 制动失效brak e flus her 制动液自动更换装置brak e shoe grind er 制动蹄片磨削装置b ranch支线bra nch li ne 岔线b rass 黄铜brazi ng byflame火焰钎焊b reakdo wn 断裂b ridgetype b akingoven 桥式烘干室b rightcoatin g 光亮镀层bright washe r 光垫圈b righte ner 光亮剂brit tlenes s 脆性br oachin g 拉削br oachin g gear拉齿bro nze 青铜bulkgoods散装货物bu rn rub ber 轮胎烧耗burn ing 烧伤burnis hing 磨光(抛光0burnis hing 抛光burr(金属)毛边flash(塑件)毛边b us bay港湾式车站bus pr iority sytem公共汽车b us pri oritysytem公共汽车优先通行系统b us she lter 候车亭bush ing bl ock 衬套busine ss law and r egulat ion 企业法规but t resi stance weldi ng 电阻对焊buttseam w elding滚对焊b utt we lding对接焊but tressthread form锯齿形螺纹buzzle蜂鸣器ca bin 客舱cablebracke t 电缆托架cadmiu m bron ze 镉青铜cage保持架cal cium c arbide电石cal culate d bend ing mo ment 计算弯矩ca m 凸轮ca m , ca m mech anism凸轮机构ca m bloc k 滑块c am dri ver 铡楔cam pr ofile实际廓线ca m prof ile 凸轮廓线cam withoscill atingfollow er 摆动从动件凸轮机构camber gauge外倾测量器camph or 樟脑c amsahf t copi ng lat he 凸轮轴仿形车床c amshaf t grin ding m achine凸轮轴磨床camsha ft pol ishing machi ne 凸轮轴超精磨机床camsha ft pol ishing machi ne 凸轮轴抛光机床ca mshaft(journ al )tu rninglathe凸轮轴(轴颈)车床can tileve r beam悬臂梁ca ntilev er str ucture悬臂结构c apabil ity 能力car r ow 车列c ar sta nd (ja ck sta nd) 汽车架carbi de towatergenera tor 投入式乙发生器carbon arc w elding碳弧焊ca rbon e lectro de 碳极carbon fouli ng (火花塞)积碳ca rbon s tructu real s teel 普通碳素结构钢carbo n tool steel碳素工具钢carbon-fiber reinf orcedplasti cs 碳纤维增强塑料c arboni tridin g (cya niding) 碳氮共渗(氰化)ca rboniz ation碳化car bureto r icin g 化油器结冰carbu rizing渗碳car burizi ng all oy ste el 渗碳合金钢car burizi ng car bon st eel 渗碳碳素钢car burizi ng fur nace 渗碳炉car d of g oods 货卡carri age 车厢Cartes ian co ordina te man ipulat or 直角坐标操作器c arton纸箱cart on box纸箱cas cade s peed c ontrol串级调速case-b ased d esign,CBD 基于实例设计ca sh far e 普通票casing headgasoli ne 气体汽油castbronze锻造青铜cast i ron 铸铁cast s teel 铸钢catal og 目录cathod e shie dl 阴极罩cathol yte 阴极电解液ca ting m achine浇注机ca tion e lectro phoret ic coa ting 阳离子型电泳涂料caus e anal ysis 原因分析cau se des cripti on 原因说明cavi tation穴蚀cel lophan e 赛璐酚c ellulo id 赛璐珞cellul ose ac etate醋酸纤维(脂)cemen t gum胶泥ceme ntite渗碳体cen ter di stance中心距ce nter d istanc e chan ge 中心距变动cen ter of mass质心cent er ofpressu re 压力中心cente rlessgrindi ng 无心磨削cent ral ge ar 中心轮centra lizedtransp ortati on 集中运输centr ifugal force离心力c entrif ugal f orce 向心力cent rifuga l seal离心密封c entrif ugal s tress离心应力c etanenumber十六烷值c hain 链chain链条chai n 链条槽c hain d ottedline 点划线cha in gea ring 链传动装置ch amfer倒角cham ois 麂皮change gearchange wheel变速齿轮charac ter di e 字模ch aracte ristic s 特性ch artere d vehi cle 包车chass is 底座c hassis基座cha ssis 基座chass is dyn amomet er 底盘测功机cha ssis l ubrica tor 底盘润滑机che ck nut锁紧螺母c hecked by 初审check-up 技术检查chem ical -heat t reatme nt 化学热处理che micalmechan ical w orking化学机械加工chem ical v aporou s depo siton化学气相沉积法chem ical w ear 化学磨损chil led ca st rio n 冷激铸铁chloro prenerubber氯丁橡胶Chroma te 铬酸处理chrom e bron ze 铬青铜chrome l 铬镍合金chuck for g rindin g hole and e nd fac e of b evel g ear 磨圆锥齿轮孔和端面用的卡盘circui t brea ker ca binet断路器柜ci rcular bendi ng die压圆模c ircula r gear圆形齿轮c ircula r pitc h 齿距ci rcular pitch; pitc h of t eeth 节距circ ular t hickne ss 圆弧齿厚circu lating power load循环功率流circul ation周转city month ly tic ket 市区月票city passe nger f low 市区客流cit y tran sposrt ation市区运输cl ass of vehic le mai ntenan ce 汽车维护类别cl ass of vehic le rep air 汽车修理类别cl assifi cation整理cl assifi cation of st andard s docu ment.标准文献分类clean er 清洗剂cleani ng clo th 抹布c leanne ss 清扫c learan ce 径向间隙cloc kwise顺时针clo gged f ilter滤清器阻塞c lose t ype so cket j oint 封闭式插接c losedchainmechan ism 闭链机构clos ed kin ematic chain闭式链c loud p oint 浊点clubcar 高尔夫球车clu tch 离合器clut ch exp losion离合器炸裂coarse threa d 粗牙螺纹coati ng for prote ctionagains t corr osion防腐镀层c ode fo pract ice 实施规则coef ficien t of a vailab ilityof veh icle 车辆完好率c oeffic ient o f fric tion 摩擦系数coe fficie nt ofspeedfluctu ation机械运转不均匀系数co effici ent of speed fluct uation速度不均匀( 波动) 系数c oeffic ient o f trav el spe ed var iation行程速度变化系数co effici ent of utili zation of au tomobi le 出车率coeff icient of ut ilizat ion of ton-k ilomet ers 吨公里利用系数coeffi cientof vel ocityfluctu ation运转不均匀系数coil stock卷料coi nciden t poin ts 重合点coinin g pres s 精压机coinin g pres s 精压机c oke-fi red fu rnace焦碳炉co ld bri ttlene ss 冷脆性cold d raw 冷拉colddrawin g lowcarbon seaml ess st eel tu be 冷拔低碳无缝钢管cold d rawing sprin g stee l 冷拉弹簧钢coldextrud ing 冷挤压cold forgi ng 冷锻c old fo rgingdie 冷锻模coldpressu re wel ding 冷压焊col d rins e bank冷水槽co ld rol led th in ste el she et 冷轧薄钢板col d roll ing ki lled s teel s heet 镇静钢冷轧钢板coldrollin g rimm ed ste el she et 沸腾钢冷钆钢板c old ru nning-in 冷磨合cold s lug 冷块cold t reatme nt 冷处理coldtrim 冷切边coll apse o f pist on ski rt 活塞裙部挤扁co llecti on ofstanda rd dec umnets标准文献收集comb inatio n in p aralle l 并联式组合combi nation in se ries 串联式组合c ombina ton pl iers 鲤鱼钳comb ined e fficie ncy; o verall effic iency总效率co mbined mecha nism 组合机构com binedstress复合应力combus ion te ste 燃烧分析仪com bustio n test e 燃烧(废气)分析仪commer cial b ronze工业用铜co mmon a pex of cone锥顶com mon eq uipmen t 常用设备common equip ment 常用设备co mmon g oods 普通货物com mon no rmal l ine 公法线comm uter 月票乘客com pany s tandar d(ente rprise stand ard) 企业标准co mparab le sta ndard可比标准co mpensa tion 补偿comp ensati on far e 补票co mpleme ntarystanda rd 补充性标准com pleteanneal ing 完全退火comp lete f ailure完全故障comple te pen etrati on 焊透c omplex mecha nism 复杂机构co mplexsteelsheetwith c hromat e zinc铬酸锌复合钢板com posite tooth form组合齿形co mpound (or c ombine d) gea r trai n 复合轮系compo und co mbinin g 复合式组合compo und di e 合模c ompoun d flat belt复合平带co mpound geartrain混合轮系c ompoun d hing e 复合铰链Compou nd scr ew mec hanism复式螺旋机构comp ressed gas 压缩煤气com pressi on str ength抗压强度c ompres sive s tress压应力com presso n coil sprin g 压缩螺旋弹簧com presso r 压缩机c ompres sor oi l 压缩机机油compu ter ai ded de sign,CAD 计算机辅助设计comput er aid ed man ufactu ring,CAM 计算机辅助制造comput er int egrate d manu factur ing sy stem,CIMS 计算机集成制造系统con cave 凸concav ity 凹面、凹度Con ceal I nstall暗装co nceptdesign, CD 方案设计、概念设计con clusio n 结论co ncurre d desi gn, CD并行设计concur rent e nginee ring 并行工程con denser teste r 电容器试验器con dition of se lf-loc king 自锁条件con ductio n of h eat 导热性cone angle圆锥角co ne dis tance锥距cone-wormmillin g mach ine 球面蜗杆铣床c onical rolle r bear ing 圆锥形滚子轴承c onical sprin g 锥形弹簧conju gate c am 共轭凸轮conju gate p rofile s 共轭齿廓conju gate y oke ra dial c am 等径凸轮conne ctingrod al ignmen t fixt ure 连杆矫正器co nnecti ng rod, coup ler 连杆connec tion b ox 接线盒connec tion t est 检查接线con oid he lical-coil c ompres sion s pring圆锥螺旋扭转弹簧con servat ion 清洁consta ntan 康铜const ant-br eadthcam 等宽凸轮con stant-veloci ty (or doubl e) uni versal joint双万向联轴节cons tituti on ofmechan ism 机构组成cons tituti on ofstanda rd 标准的构成con strain ing fo rce 约束反力cons traint约束con strain t cond ition约束条件c onsump tion 消耗conta ct poi nts 啮合点conta ct rat io 重合度contac t seal接触式密封conta ct str ess 接触应力cont ainertransp ort 集装箱运输co ntent目次cont ent of norma tive d ocumen t 标准文件的内容co ntinou s furn ace 连续炉conti nuousbroach ing 连续拉削con trol c enter调度中心co ntrolpanel控制屏con trol s tation调度站C ontrol switc hes 控制开关conv ention al mec hanism; mech anismin com mon us e 常用机构conve x 凹con vex 凸的,凸面体co nvex r oller球面滚子Co nveyer流水线物料板Conv eyer 流水线物料板c oolant冷却液co ordina te 座标c oordin ate fr ame 坐标系copp er 紫铜C oppercore p ower c able 铜芯电力电缆c opy gr inding mahci ne 靠模磨床copy lathe仿形车床c ork 软木correc ting p lane 平衡平面cor rectin g plan e 校正平面corri genda勘误表cor rosion inhib itor 搞腐蚀剂cor rosion resis tance耐腐蚀性c orrosi on res isting castiron 耐蚀铸铁cor rosion wear腐蚀性磨损cosine accel eratio n (orsimple harmo nic) m otion余弦加速度运动cosm etic d efect外观不良co smetic inspe ct 外观检查cosme tic in spect外观检查c ost 成本cost o f labo r 人工费c oulome ter 库仑计count er sin king 锪孔coun terclo ckwise (or a nticlo ckwise) 逆时针c ounter weight平衡重c ouple力偶coup ler-cu rve 连杆曲线coup ling s haft c ouplin g 联轴器coverplate盖板cove red ar c weld ing 手工电弧焊co vering power深镀能力c rank 曲柄crank angle betwe en ext reme (or lim iting) posit ions 极位夹角cra nk arm, plan et car rier 系杆crank press曲轴压力机crank shaft曲柄轴cr ank sh aft 曲轴crankshaper (guid e-bar) mecha nism 曲柄导杆机构crank-rocker mecha nism 曲柄摇杆机构c ranksa hft jo urnalgrindi ng mac hine 曲轴主轴颈磨床crank shaft(mainjourna l turn ing la the) 曲轴(主轴颈)车床cra nkshaf t dyna mic ba lancin g mach ine 曲轴动平衡机c ranksh aft gr inding machi ne 曲轴磨床cran kshaft journ al fac e hard eningmachin e 曲轴轴颈表面淬火机床crank shaftjourna l mill ing ma chine曲轴轴颈铣削专用机床c ranksh aft po lishin g mach ine 曲轴轴颈抛光机床crank shaftsuperf inishi ng mac hine 曲轴超精磨机床crank shaftturnin g lath e 曲轴车床creati on des ign 创新设计cre pon fi nish 皱纹漆crim p pape r 瓦楞纸c ritica l defe ct 极严重缺陷cri ticalfailur e 致命故障critic al poi nt 临界点critic al spe ed 临界转速cros s grin ding 端面磨削cro ss wel d 横向焊缝cross-belt d rive 交叉带传动c rossed helic al gea rs 交错轴斜齿轮cro ssover交叉器cr oss-sh aped j oint 十字接头cr own ge ar 冠轮c rown s having剃鼓形齿c rusher破碎机c ryptom eter 遮盖力测定仪c rystal lizati on poi nt 结晶温度CSBS yearb ook 中国标准化年鉴c st bim etal 铸造双金属c ulture教养cup ola (c upolafurnac e) 冲天炉curb w indow安全监视窗curlin g 卷边Cu rrentby Pha se 每相电流curre nt rep air of vehic le 汽车小修curv ature曲率curv e matc hing 曲线拼接cur ved-sh oe fol lower曲面从动件curve-toothbevelgear m illing弧齿锥齿轮铣刀盘刃c urve-t ooth b evel g ear mi llingmachin e 弧齿锥齿轮铣齿机c urvili near m otion曲线运动c ushion缓冲cut-off (shear-out) 切口cutte r 刀具cu tter g rindin g mach ine 工具磨床cut ting m achine切割机cu tting-out pr ess 切断压力机cya niding氰化cy cle 周波cycleof mot ion 运动周期cycl oidalgear 摆线齿轮cy cloida l moti on 摆线运动规律cyc loidal tooth profi le 摆线齿形cycl oidal-pin wh eel 摆线针轮cyli nder b lock b roachi ng mac hine 气缸体专用拉床cylin der bl ock si de bro aching machi ne 气缸体侧拉床cy linder borin g mach ine 镗缸机cyli nder c ompres sion g auge 气缸压力表c ylinde r honi ng mac hine 气缸珩磨机c ylinde r leak teste r 气缸漏气率检验仪c ylinde r perp endicu larity gauge气缸孔垂直检验仪cy linder press ure ga uge 气缸压力表cy linder score拉缸cyl indersticki ng 咬缸cylind ric pa ir 圆柱副cylind ricalcam 圆柱凸轮cyl indric al coo rdinat e mani pulato r 圆柱坐标操作器cy lindri cal gr inding外圆磨削c ylindr ical r oller圆柱滚子c ylindr ical r ollerbearin g 圆柱滚子轴承cyli ndrica l worm圆柱蜗杆cylind roid h elical-coilcompre ssionspring圆柱螺旋压缩弹簧cy lindro id hel ical-c oil ex tensio n spri ng 圆柱螺旋拉伸弹簧cylind roid h elical-coiltorsio n spri ng 圆柱螺旋扭转弹簧D.C tr action subst ation直流牵引变电站D.I. rinse纯水次da ily se vice 每日保养da mage 损坏dateof sta ndardimplem entati on 标准实施日期da ted re ferenc e (tostanda rds) 注明日期的引用(标准)d ay and night line昼夜线路da y andnightvehicl e 昼夜车dead p oint 死点deade ner 防声胶decan tation沉淀法d eceler ated r un 压点d ecentr alized trans portat ion 分散运输dec isionitems决议事项de coilin g unit开卷机de dendum齿根高。
Two-dimensional Gross-Neveu Model with Wilson Fermion Action at Finite Temperature and Dens
a r X i v:h ep -l a t /9809071v 1 11 S e p 19981UTHEP-387UTCCP-P-47Two-dimensional Gross-Neveu Model with Wilson Fermion Action at Finite Temperature and Density ∗T.Izubuchi a ,J.Noaki a and awaa ,baInstitute of Physics,University of Tsukuba,Tsukuba,Ibaraki 305-8571,JapanbCenter for Computational Physics,University of Tsukuba,Tsukuba,Ibaraki 305-8577,JapanWe analytically investigate the 2-dimensional Gross-Neveu model at finite temperature and density using Wilson fermion action.The relation between the phase structure on the lattice and that in the continuum is clarified.1.INTRODUCTIONThe 2-dimensional Gross-Neveu model[1]has often been used to examine theoretical issues of QCD.This stems from the fact that the model not only shares the property of asymptotic free-dom and spontaneous breaking of chiral symme-try with QCD,but also can be solved analytically in 1/N expansion.In this article we present a summary of an yet another study[2]of this cat-egory.The problem addressed is how the phase diagram of the continuum theory at finite tem-perature and density[3]emerges from that of the lattice theory formulated with the Wilson fermion action.In particular we explore how this limit is achieved in the presence of the parity-broken phase due to the Wilson action[4,5].An addi-tional interest is to gain an understanding on how the tricritical point of the continuum model,sep-arating a first-and second-order chiral transition on the (T,µ)plane,emerges from the lattice point of view.TTICE GROSS-NEVEU MODEL Our lattice model is given by the Lagrangian,L =12a ¯ψn γ2(e −µa ψn +ˆ2−e µa ψn −ˆ2)+δm ¯ψn ψn −12¯ψn ∇2µψn (1)g 2π=13/27+1/(12π).Therefore wehave to study the lattice phase structure in the3-dimensional space spanned by (g 2σ,g 2π,δma )TTICE PHASE STRUCTURES AND THE CONTINUUM LIMIT We analyze the phase structure in terms of the effective potential V (σ,π)and the saddle-point equations ∂V/∂σ=0,∂V/∂π=0calculated to leading order in 1/N .The latter equations exhibit both parity-broken solution π=0and symmetric solution π=0.One may expect that the phase boundary between the two solutions is determined by taking the limit π→0from the parity-broken phase,in which case the transition230.00.10.20.30.40.50.60.71.51.00.50.0N T =200N T =500N T =1000continuumFigure 2.σ/Λas a function of µ/Λat at T/Λ=0.2for various temporal size N T together with the continuum result.0.00.20.40.80.60.40.20.0N T =100N T =200N T =500N T =1000continuumFigure 3.Same as Fig.4for T/Λ=0.5.σ>0.The new phase is characterized by a small value σ≈0,therefore chiral symmetry is effec-tively restored in it.The two phase boundaries JKQP and LMQP are both of first order.Hence,when the curve with arrow defining the contin-uum limit penetrates the boundary surface,as illustrated in Fig.1(c),chiral symmetry becomes restored through a first-order phase transition.3.4.T =0and µ=0The phase diagram and order of chiral phase transition for general values of temperature and chemical potential are determined by a combina-tion of the finite-temperature and finite-density effects discussed above.At relatively low temper-atures,the situation as a function of µis similar to that at T =0.Hence a first-order chiral transi-tion occurs.On the other hand,at high tempera-tures,the behavior of phase structure is the same type as that at µ=0,leading to a second-order phase transition.Let us define a Λ-parameter by Λ=c ·e −π/g 2σ/a (c =0.5716···)with which σ/Λ=1at T =µ=0in the continuum limit.In Fig.2and Fig.3,we plot σ/Λas a function of µ/Λfor a set of values of N T at T/Λ=0.2and 0.5.We observe a clear first-order transition for the low-temperature case (Fig.2),which changes to a second-order transi-tion at high temperatures (Fig.3).Furthermore,the curves for finite N T converge to the contin-uum result (thick solid line)at N T →∞.These results show how the lattice phase structure leads to the phase diagram of the continuum theory with a tricritical point[3].4.SUMMARYThe phase structure of QCD in the (T,µ)plane is still largely unknown,and various interesting possibilities have recently been discussed[8].Lat-tice study of these possibilities with the Wilson quark action has to disentangle effects of parity-broken phase from physical ones.Our results should be helpful in the effort in this direction.This work is supported in part by Grants-in-Aid of the Ministry of Education (Nos.2375and 10640246),and by the JSPS Research for Future Program.TI is a JSPS Research Fellow.REFERENCES1. D.J.Gross and A.Neveu,Phys.Rev.D 10(1974)3235.2.T.Izubuchi,J.Noaki and awa,hep-lat/9805019,to appear in Phys.Rev.D.3.U.Wolff,Phys.Lett.157B (1985)303.4.S.Aoki,Phys.Rev.D 30(1984)2653.5.S.Aoki,awa and T.Umemura,Phys.Rev.Lett.76(1996)873.6.T.Eguchi and R.Nakayama,Phys.Lett.126B (1983)89.7.S.Aoki and K.Higashijima,Prog.Theo.Phys.76(1986)521.8.See,e.g.,M.Alford,in these proceedings.。
基于自适应多尺度超螺旋算法的无人机集群姿态同步控制
基于自适应多尺度超螺旋算法的无人机集群姿态同步控制蔡运颂 1, 2许 璟 1, 2牛玉刚1, 2摘 要 四旋翼无人机(Unmanned aerial vehicle, UAV)系统姿态角和角速度分别为运行在不同时间尺度上的慢、快动态. 由于输入扰动的上界难以精确估计, 本文提出一种基于自适应多尺度超螺旋(Super-twisting, STW)滑模算法的无人机集群一致性控制策略. 首先, 建立无人机集群系统的姿态角模型, 并通过奇异摄动理论将其化为两时间尺度形式. 基于系统的快慢特性, 本文设计两时间尺度的超螺旋滑模算法, 并采用自适应增益处理无人机集群系统的未知边界非线性. 此外,还提出一种改进型自适应多尺度超螺旋滑模算法, 进一步减少系统的一致性收敛时间, 实现无人机集群姿态角有限时间内同步. 最后通过仿真分析, 验证两种自适应多尺度超螺旋算法的正确性和有效性.关键词 奇异摄动, 超螺旋算法, 多尺度, 姿态协同, 四旋翼无人机引用格式 蔡运颂, 许璟, 牛玉刚. 基于自适应多尺度超螺旋算法的无人机集群姿态同步控制. 自动化学报, 2023, 49(8):1656−1666DOI 10.16383/j.aas.c220759Attitude Consensus Control of UAV Swarm Based onAdaptive Multi-scale Super-twisting AlgorithmCAI Yun-Song 1, 2 XU Jing 1, 2 NIU Yu-Gang 1, 2Abstract In a UAV (unmanned aerial vehicle) system, the attitude angle and angular velocity of the UAV are, re-spectively, the slow and fast dynamics operating in different time scales. Due to the difficulty in the estimation of the bound of disturbance, this paper proposes a control method for UAV swarm, based on the adaptive multi-scale STW (super-twisting) sliding mode algorithm. First, the attitude model of the UAV swarm system is established,which is transformed into a two-time-scale model via singular perturbation theory. On this basis, this paper designs a two-time-scale STW sliding mode algorithm with adaptive gains to deal with the perturbations and unknows.Furthermore, by adding a few linear iterms, a modified adaptive STW control algorithm is also provided, which further reduces the convergence time and achieves the synchronization of the attitudes in finite time. Finally,the effectiveness of two different adaptive multi-scale STW algorithms are verified through simulations.Key words Singular perturbation, STW, multi-scale, attitude coordination, quadrotorsCitation Cai Yun-Song, Xu Jing, Niu Yu-Gang. Attitude consensus control of UAV swarm based on adaptive multi-scale super-twisting algorithm. Acta Automatica Sinica , 2023, 49(8): 1656−1666四旋翼无人机[1−2](Unmanned aerial vehicle,UAV)具有结构简单、飞行精准、机动性强等优点.因此, 在军事打击[3−4]、载物[5−6]、测量[7−8]、灾害监测[9]等方面, 有着很好的应用. 然而随着控制任务复杂度的增加, 例如无人机表演[10]、沿海侦察、集群打击等, 仅凭一台无人机难以完成, 因此需要多台无人机集群协同作业. 在对四旋翼无人机进行建模时,通常简单地认为无人机模型是单一尺度的. 然而实际上, 无人机的姿态角与角速度并不处于同一时间尺度, 这是由无人机中的参数量纲差异引起的. 因此, 无人机集群的奇异摄动建模具有重要意义, 通过奇异摄动建模可以抽提出无人机状态的快慢特性. 然而, 对于奇异摄动无人机集群系统而言, 基于单一时间尺度的控制策略效果欠佳.目前, 四旋翼无人机集群控制方法主要有反步法、模糊控制以及PID (Proportion-integral-deriv-ative)控制方法等. 文献[11]针对多四旋翼无人机的编队控制, 采用反步法实现了四旋翼无人机群对期望轨迹的跟踪功能. 文献[12]建立了四旋翼无人机的姿态动力学模糊模型, 设计了模糊反馈控制器,收稿日期 2022-09-22 录用日期 2023-02-10Manuscript received September 22, 2022; accepted February 10, 2023国家自然科学基金(62173141, 62073139), 上海市自然科学基金(22ZR1417900)资助Supported by National Natural Science Foundation of China (62173141, 62073139) and the Natural Science Foundation of Shanghai (22ZR1417900)本文责任编委 李鸿一Recommended by Associate Editor LI Hong-Yi1. 华东理工大学信息科学与工程学院 上海 2002372. 华东理工大学能源化工过程智能制造教育部重点实验室 上海 2002371. College of Information Science and Engineering, East China University of Science and Technology, Shanghai 2002372. Key Laboratory of Intelligent Manufacturing of Energy and Chemical Processes of Ministry of Education, East China University of Sci-ence and Technology, Shanghai 200237第 49 卷 第 8 期自 动 化 学 报Vol. 49, No. 82023 年 8 月ACTA AUTOMATICA SINICAAugust, 2023实现了四旋翼无人机集群控制. 文献[13]设计了一种BP (Back propagation)神经网络辅助的PID 无人机编队智能算法, 实现了PID 参数的优化整定[14].对四旋翼无人机集群的研究中, 姿态协同是四旋翼无人机群实现队形控制、协同避障等任务的基础.文献[15]基于半定规划进行迭代区域扩张完成了多无人机的队形设计. 文献[16]利用神经网络预测姿态偏差, 将其集成于分散式容错协同控制器中,实现了姿态角的一致性. 然而, 考虑到无人机的动态模型中存在着内部结构不确定, 外界扰动影响等问题, 导致基于无扰动简化模型的控制方案效果有限.sgn (·)在姿态协同控制中, 滑模控制是一类有效的鲁棒控制方法, 对于外部输入扰动或者参数不确定性具有不变性、有限时间可达等优点. 目前, 滑模控制方法大致可以分为一阶滑模与高阶滑模. 如文献[17]基于一阶滑模与低通滤波器的结合, 实现了对直流电机位置的控制. 然而, 一阶滑模是直接基于滑模变量的一阶导数设计的, 采用了切换控制律, 产生了严重的抖振现象, 影响系统性能. 在二阶滑模算法中, 超螺旋滑模算法(Super-twisting, STW)的应用最为广泛. 这是由于超螺旋滑模控制器采用了连续控制结构, 引入了积分项, 避免了使用切换项, 响应速度快, 对抖振抑制能力强, 并且可以驱使滑模变量及其导数在有限时间内收敛到稳定点. 同时, 能够处理上界为依赖于状态的函数以及符合Lipschitz 条件[18]的扰动. 文献[19]采用了超螺旋滑模控制策略, 提高了永磁同步电机的转速响应. 文献[20]提出了一种基于超螺旋滑模的干扰观测器, 实现了对未估计的干扰的精细化补偿. 然而上述滑模控制方法都是基于已知上界的非线性, 这在无人机中是无法实现的.为了实现姿态协同的稳准快, 本文设计了一种新型的分尺度自适应STW 算法, 通过分尺度自适应STW 控制器产生的不同时间尺度上的快、慢控制律, 实现了四旋翼无人机奇异摄动多智能体模型中的分尺度精确控制. 同时, 通过自适应增益实现扰动未知情况下的快速补偿. 与现有部分研究成果相比, 本文的主要贡献归纳为如下几个方面:1) 多时间尺度超螺旋控制结构: 本文提出了多时间尺度超螺旋滑模控制器的设计方法, 在控制器中引入两个时间尺度, 通过奇异摄动方法来有效处理四旋翼无人机姿态角系统状态同步问题.2) 自适应分布式控制器: 本文采用了分布式的控制结构, 对每个四旋翼无人机智能体分别设计了一个自适应增益, 让其自适应于四旋翼无人机智能体本身以及与其他智能体间的耦合.n ×n X −1T ⊗符号描述. 对于一个 维的矩阵 , 上标 表示矩阵的逆, 上标 表示矩阵的转置, 表示diag {a 1,a 2,a 3}a 1,a 2,a 3n b T sgn col {b 1,b 2,b 3}|b |b ||b ||b 12b min (ω)ωmax (ω)ωI 303×03×1×矩阵的克罗内克积, 表示对角线上的元素为 的矩阵. 对于一个 维向量 ,上标 表示向量的转置, 表示符号函数, 表示向量按列排序, 表示 内元素取绝对值后的向量, 表示向量的二范数, 表示 内元素开根号后的向量. 表示取集合 中最小的数, 表示取集合 中最大的数. 与 分别表示3 3的单位对角阵与零矩阵. 表示3 1的零矩阵.1 四旋翼无人机模型n 假设四旋翼无人机多智能体系统中具有 个四旋翼无人机智能体, 单个四旋翼无人机的姿态非线性动力学方程为[21]:i =1,···,n,ϕi θi ψi i ϕi ∈(−π/2,π/2)θi ∈(−π/2,π/2)ψi ∈(0,2π)I xi ,I yi ,I zi x b y b z b J ri w ri =w 1i −w 2i +w 3i −w 4i w 1i ,w 2i ,w 3i ,w 4iJ ri w ri ˙θi J ri w ri ˙ϕi k axi ,k ayi ,k azi u 1i ,u 2i ,u 3i 其中, 、 、 分别表示第 个无人机的横滚角、俯仰角、偏航角, 、 、 , 表示无人机体绕机体坐标系 , , 轴的转动惯量, 表示无人机的电动机和桨叶的转动惯量. 输入扰动为, 其中, 表示无人机四个旋翼的转速, 、 表示陀螺力矩, 表示空气阻力矩系数, 表示无人机旋翼对其三个姿态角的控制量.I xi ,I yi ,I ziϵ=min (I xi ,I yi ,I zi )¯Ixi =I xi /ϵ¯I yi =I yi /ϵ¯I zi =I zi /ϵ¯J ri =J ri /ϵ¯kaxi =k axi /ϵ¯k ayi =k ayi /ϵ¯k azi =k azi /ϵˆI i =diag {¯I xi ,¯I yi ,¯I zi }xi =(ϕi ,θi ,ψi )T vi =(˙ϕi ,˙θi ,˙ψi )T u i =ˆI −1i(u 1i ,u 2i ,u 3i )T i 由于四旋翼无人机存在着小参量 等, 呈现较为显著的奇异摄动现象[22]. 因此对四旋翼无人机智能体系统进行奇异摄动的建模. 定义, , , , , , ,, , , , . 基于此, 第 个四旋翼无人机的姿态非线性动力学矩8 期蔡运颂等: 基于自适应多尺度超螺旋算法的无人机集群姿态同步控制16572 系统描述与引理将式(2)表示为状态空间方程:G =[a ij ]a ij i j i,j =1,2,3,···,m 假设每个智能体都可以访问邻接的智能体的输出相对值, 并且相关的邻接矩阵表示为 , 其中 表示第 个智能体与第 个智能体之间连接的权值, 若无连接则为0, 且 .定义一致性角度误差和角速度误差为:由式(3)、(4), 可得以下同步误差模型:为了后续分析, 在此给出假设和引理.¯φi (t,g i )=∑n j =1,j =i a ij (g i (v i ,w ri )−g j (v j ,w rj ))||¯φi (t,g i )||≤δi ||s i (t )||12¯φi (t,g i )δi >0假设 1. 令 , 且 , 其中, 满足Lipschitz 条件,存在但未知.Z i (i =1,···,5)Z i =Z T i (i =1,···,4)引理 1[23]. 若存在矩阵 且 , 满足以下线性矩阵不等式:Z (ϵ)>0ϵ∈(0,¯ϵ]Z (ϵ)=[Z 1+ϵZ 3ϵZ T5ϵZ 5ϵZ 2].则可以得到 , 对任意的 都成立, 其中, z 1z 2z 3引理 2. 对于任意列向量 , 和 . 有以下不等式成立:a =z 2z T2z 3b =z 1证明. 令 , , 则l >0x y ±xy <lx 2+14l y 2引理 3[24]. 给定任意正定标量 , 对于任意标量 , , 有以下不等式成立: .n x P n ×n 引理 4[25]. 对于一个 维非0列向量, 为 维的Hermitian 矩阵, 有如下性质:3 四旋翼无人机集群姿态角一致性分析设计以下受导引型奇异摄动二阶滑模动态:l i l 1>0l i =0(i =1)x 0(t )s i (t )其中, 表示追踪系数, , , 表示姿态角的跟踪值, 表示第i 个滑模变量.根据式(5)可得, 滑模动态(6)的一阶导数为:3.1 自适应多尺度超螺旋算法受文献[26]的启发, 考虑到系统(5)的两时间尺度特性, 设计以下自适应STW 滑模控制器:αi (t )βi (t )其中, 和 表示两个自适应增益.将式(8)代入式(7), 可得:下面的定理研究了在自适应多尺度STW 算法控制下的四旋翼无人机群在有限时间内的姿态协同.p 1i >0p 2i >0b 1i >0b 2i >0γ1i >0γ2i >0¯ϵ>0定理 1. 给定 , , , ,, , 存在 , 当满足:1658自 动 化 学 报49 卷以及系统的自适应增益导数满足:ϵ∈(0,¯ϵ]则对任意的 , 四旋翼无人机集群系统的姿态角将会在有限时间内趋于一致.证明. 构造新的状态变量:根据式(9)、(12), 可得:ηi =col {−z 1i z T1i2||z 1i ||3(z 2i +φi (t,g i )),03×1}φi (t,g i )=diag {sgn (s i (t ))}¯φi (t,g i )其中, , .z 1i z 2i s i˙si 由式(11)、(13), 可知: 当 , 趋于0时, 会趋于0, 再根据式(9)以及假设1, 也会趋于0.考虑以下奇异摄动Lyapunov 函数:F ϵ=diag {1,ϵ}其中, ˆαi αi (t )ˆβi βi (t )P i (ϵ)>0V 0i (t,ϵ)=z TP i (ϵ)z i 表示 的上界, 表示 的上界. 根据引理1, 成立的充分条件为式(10). 定义, 并对其求导可得:由假设1, 可知:可以构造以下不等式:其中,p (t )=2z T i ¯P i (ϵ)ηi 令 , 易得:由引理2, 将式(16)转化为:由引理3, 可构造:联立式(17)和式(18), 可得:Y i (ϵ)=diag {d 1,d 2}其中, 联立式(14)、(15)、(19), 可得:8 期蔡运颂等: 基于自适应多尺度超螺旋算法的无人机集群姿态同步控制1659βi (t )=−ϵp 2i 2p 3iαi (t )+p 1ip 3i Q i (ϵ)>0设计 . 根据Schur 补引理[27], 可得 , 当以下条件成立时:由引理4, 可知:基于式(21), 我们有:根据式(20)、(21)、(22), 可得:r 1i =λmin (Q i )λ12min (P i)λmax (P i )其中, .βi (t )≤ˆβi αi (t )≤ˆαi 由于 , . 结合式(23), 可得:1γ1i˙αi (t )−b 1i √2γ1i=01γ2i˙βi (t )−b 2i √2γ2i=0αi (t )βi (t )令式(24)中 , , 则可得 , 应满足式(11). 将式(11)代入式(24)中, 根据柯西不等式[28], 可得:ˇr i =min (r 1i ,b 1i ,b 2i )ˇr k =min (ˇr i )其中, , .由此可见, 四旋翼无人机集群系统的一致性误差在有限时间内稳定. □3.2 改进型自适应多尺度超螺旋算法由于定理1中, 在自适应多尺度STW 算法控制下的系统收敛时间相对较长. 因此在文献[29]的启发下, 设计以下改进型自适应STW 滑模控制器:k 1i k 2i 其中, 、 为两个增益.将式(26)代入式(6), 可得:下面的定理研究了在改进型自适应多尺度STW滑模算法的控制下, 四旋翼无人机集群系统的姿态角能够快速地趋于一致.b 3i >0b 4i >0b 5i >ˆαi b 6i >ˆβi ¯ϵ>0定理 2. 给定 , , , . 在控制器(26)作用下, 系统状态将快速趋于一致, 当存在 , 使得以下式子成立时:其中,对应的相关参数为:1660自 动 化 学 报49 卷证明. 构造新的状态变量:根据式(27)、(29), 可得:E ϵ=diag {I 3,I 3,ϵI 3}其中,考虑以下奇异摄动Lyapunov 函数:P 2i (ϵ)=(¯P2i (ϵ)E ϵ)⊗I 3>0其中, ,V 20i (t,ϵ)=ˆz T i P 2i (ϵ)ˆz i 定义 ,对其求导可得:对应的相关参数为:由假设1, 可知:可以构造以下不等式:其中,8 期蔡运颂等: 基于自适应多尺度超螺旋算法的无人机集群姿态同步控制1661p 2(t )=2ˆz T i ¯P 2i (ϵ)η2i 令 , 易得:根据引理2, 将式(33)转化为:根据引理3, 可构造:联立式(34)、(35), 可得:d 4=p 11i 4c +p 12i 4c −ϵp 13i (1+c 8i )其中, ,联立式(31)、(32)、(36), 可得:其中,W i (ϵ)>0X i (ϵ)>0˙V20i (t,ϵ)<0由式(37)可知: , 时,成立.由引理4, 可知:基于式(38), 可得:根据式(37)、(39), 可得:r 2i =λ12min (P 2i )λmin (¯W i )λmax (P 2i ),r 3i =λmin (¯Xi )λmax (P 2i )其中, .将式(40)代入式(30), 可得:根据柯西不等式[28], 将式(41)转化为:1662自 动 化 学 报49 卷ˇr 2i =min (r 2i ,b 1i ,b 2i )ˇr 3i =min (r 3i ,b 3i ,b 4i )其中, , .1γ1i˙αi (t )−b 3i b 5i2γ1i−b 1i √2γ1i=01γ2i ˙βi (t )−b 4ib 6i2γ2i−b 2i √2γ2i=0令式(42)中 ,, 可得:则式(42)可转化为:结合式(30)、(43), 根据柯西不等式[28], 可得:ˇr 2k =min (ˇr 2i )ˇr 3k =min (ˇr 3i )其中, , .即在改进型自适应STW 滑模控制器(26)的作用下, 无人机集群系统的误差有限时间内稳定.□4 一致性误差收敛时间分析在本节, 我们将比较自适应多尺度STW 算法和改进型自适应多尺度STW 算法的收敛时间, 进一步分析改进型算法具有更短的收敛时间的原因.在控制器(8)的作用下, 根据式(25), 可得:z i t r 1d t V12(t,ϵ)[t 0,t r 1]假定状态 在 时刻收敛, 将式(45)两边同乘, 并在 上进行积分:t 0=0z i t r 1V 12(t r 1,ϵ)=0其中, , 状态 在 时刻收敛, 即 , 代入式(46)可得:在控制器(26)的作用下, 由式(44)可得:ˆzi t r 2d tˇr 2k ¯V12(t,ϵ)+ˇr 3k ¯V(t,ϵ)[t 0,t r 2]假定状态 在 时刻收敛, 将式(48)两边同乘, 并在 上进行积分:t 0=0,f (t )=2ˇr 3k ln (1+ˇr 3k ¯V12(t,ϵ)ˇr 2k)ˆz i t r 2¯V(t r 2,ϵ)=0ln (1+ˇr 3k ¯V12(t r 2,ϵ)ˇr 2k)=0其中, . 状态 在 时刻收敛, 即 , 则 , 代入式(49), 可得:ln (·)ln (·)t r 2<t r 1由式(47)、(50)可知, 在改进型自适应STW 滑模控制器(26)的作用下, 收敛时间与 函数相关联, 由于 函数的取对数特性, 使得无人机集群系统的一致性收敛时间更短, 即 .5 仿真分析a 12=1a 13=2a 23=3为验证所建立模型与控制律的有效性, 本次仿真选择了三个四旋翼无人机智能体集群, 三个智能体之间的交互关系由无向图表示, 且 , , . 无人机间的连接方式可参照图1.132...u 1u 2u iu 3图 1 四旋翼无人机多智能体Fig. 1 The multi-agent of quadrotorsI x 1=I y 1=6.22×10−3kg ·m 2I z 1=1.12×10−3kg ·m 2I x 2=I y 2=l 9.22×10−3kg ·m 2I z 2=2.12×10−3kg ·m 2I x 3=I y 3=3.22×10−3kg ·m 2I z 3=7.12×10−4kg ·m 2J r 1=6×10−5kg ·m 2J r 2=9×10−5kg ·m 2J r 3=3×10−5kg ·m 2k ax 1=k ay 1=k az 1=1.2×10−4N ·s /m k ax 2=k ay 2=k az 2=2.2×10−4N ·s /m kax 3=k ay 3=k az 3=7.2×10−5N ·s /m ,ϵ=7.12×10−4.四旋翼无人机绕机体坐标系的转动惯量为: , ,, , , . 四旋翼无人机的电动机和桨叶的转动惯量为: , ,. 四旋翼无人机的空气摩擦阻力矩系数为: ,, 取 四旋翼无人机初始姿态角与角速度为:8 期蔡运颂等: 基于自适应多尺度超螺旋算法的无人机集群姿态同步控制1663l 1=1w ri =5sin (t )x 0(t )=(π4sin (t ),π4sin (t ),π4sin (t )+π2)T跟踪系数 , 非线性项中 , 跟踪姿态角 .为了实现无人机的姿态角的同步, 仿真中采用了两种控制器对无人机姿态集群系统进行控制:b 11=2b 12=2.2b 13=2.4b 21=1b22=1.2b 23=1.4γ11=2γ12=3γ13=4p 11=−p 22=p 31=1p 12=−p 22=p 32=1.2p 13=−p 23=p 33=1.4βi (t )αi (t )1) 采用自适应多尺度STW 控制器(8), 对应的控制器相关参数为: , , ,, , , , ,. , , . 自适应增益 , 形式如式(11)所示.k 11=1k 12=1.1k 13=1.2k 21=2k 22=2.1k 23=2.2b 3i =0.1b 4i =0.1b 5i =8b 6i =8b 21=1b 22=1.2b 23=1.4γ21=1γ22=2γ23=3βi (t )αi (t )2) 采用改进型自适应多尺度STW 控制器(26), 对应的控制器相关参数为: , ,, , , . ,, , . , , , , , . 自适应增益 , 为:图2为在自适应多尺度STW 控制器(8)作用下的四旋翼无人机的姿态角状态轨迹曲线. 从中可以看出无人机集群系统的姿态角在有限时间内实现状态同步. 图2(d)为自适应增益变化曲线, 可以看出, 自适应增益持续增加直至无人机姿态角协同.图3表明: 在改进型自适应多尺度STW 控制器(26)作用下, 也能够使得无人机集群系统姿态角在有限时间内达到一致. 两种控制算法下系统的性能指标如表1所示, 主要从平均收敛时间、平均稳态误差这两个指标进行比较. 由表1可知, 在改进型自适应多尺度STW 算法控制下的无人机集群系统的快速性明显增加, 准确性略微减弱. 相对于文献[30]提出的控制算法, 本文提出的这两种算法在收敛时间上更短, 控制的准确性更高.6 结论本文针对四旋翼无人机系统中具有的多时间尺度特性, 以及存在未知边界非线性的问题, 设计了一种自适应多尺度STW 滑模算法.将无人机快慢系统“分而治之”, 实现了分尺度精确控制. 并且通过该算法在有效削减滑模动态抖振的同时, 还保证了无人机集群系统在有限时间内的一致性. 本文还t /s−−R o l l a n g l e /r a d(a) 无人机横滚角变化曲线(a) Quadrotors roll anglechange curvet /s(b) 无人机俯仰角变化曲线(b) Quadrotors pitch anglechange curve t /s(c) 无人机偏航角变化曲线(c) Quadrotors yaw anglechange curvet /s (d) 自适应增益 a i (t ) 变化曲线(d) Adaptive gain a i (t )variation curve −−Y a w a n g l e /r adA d a p t i v e g a i n图 2 自适应多尺度STW算法控制下的无人机姿态历时曲线Fig. 2 Trajectories of attitudes under the adaptivemulti-scale STW controllert /s(a) 无人机横滚角变化曲线(a) Quadrotors roll anglechange curvet /s(b) 无人机俯仰角变化曲线(b) Quadrotors pitch anglechange curve t /s(c) 无人机偏航角变化曲线(c) Quadrotors yaw anglechange curve t /s (d) 自适应增益 a i (t ) 变化曲线(d) Adaptive gain a i (t )variation curve−−−R o l l a n g l e /r a d−Y a w a n g l e /r a d A d a p t i v e g a i n图 3 改进型自适应多尺度STW 算法控制下的无人机姿态历时曲线Fig. 3 Trajectories of attitudes under the modifiedadaptive multi-scale STW controller1664自 动 化 学 报49 卷设计了一种改进型自适应多尺度STW 滑模算法,增加了系统的快速性. 最后通过仿真验证了两种控制方法的有效性, 实现了无人机集群系统的姿态协同.ReferencesXu J, Fridman E, Fridman L, Niu Y G. Static sliding mode con-trol of systems with arbitrary relative degree by using artificial delays. IEEE Transactions on Automatic Control , 2020, 65(12):5464−54711Xu Jing, Cai Chen-Xiao, Li Yong-Qi, Zou Yun. Dual-loop path tracking and control for quad-rotor miniature unmanned aerial ve-hicles. Control Theory & Applications , 2015, 32(10): 1335−1342(许璟, 蔡晨晓, 李勇奇, 邹云. 小型四旋翼无人机双闭环轨迹跟踪与控制. 控制理论与应用, 2015, 32(10): 1335−1342)2Zhou Xiao-Cheng, Yan Jian-Gang, Xie Yu-Peng, Zhai Hong-Jun. Task distributed algorithmic for multi-UAV based on auc-tion mechanism. Journal of Naval Aeronautical and Astronautic-al University , 2012, 27(3): 308−312(周小程, 严建钢, 谢宇鹏, 翟鸿君. 多无人机对地攻击任务分配算法. 海军航空工程学院学报, 2012, 27(3): 308−312)3Chang Yi-Zhe, Li Zhan-Wu, Yang Hai-Yan, Luo Wei-Ping, Xu An. A decision-making for multiple target attack based on char-acteristic of future long-range cooperative air combat. Fire Con-trol & Command Control , 2015, 40(6): 36−40(常一哲, 李战武, 杨海燕, 罗卫平, 徐安. 未来中远距协同空战多目标攻击决策研究. 火力与指挥控制, 2015, 40(6): 36−40)4Luo C, Yu L J, Ren P. A vision-aided approach to perching a bioinspired unmanned aerial vehicle. IEEE Transactions on In-dustrial Electronics , 2018, 65(5): 3976−39845De Castro A I, Torres-Sanchez J, Pena J M, Jimenez-Brenes F M, Csillik O, Lopez-Granados F. An automatic random forest-OBIA algorithm for early weed mapping between and within crop rows using UAV imagery. Remote Sensing , 2018, 10(2):Article No. 2856Kim B O, Yun K H, Chang T S, Bahk J J, Kim S P. A prelim-inary study on UAV photogrammetry for the hyanho coast near the military reservation zone, eastern coast of Korea. Ocean and Polar Research , 2017, 39(2): 159−1687Wang Ning, Wang Yong. Fuzzy uncertainty observer based ad-aptive dynamic surface control for trajectory tracking of a quad-rotor. Acta Automatica Sinica , 2018, 44(4): 685−695(王宁, 王永. 基于模糊不确定观测器的四旋翼飞行器自适应动态面轨迹跟踪控制. 自动化学报, 2018, 44(4): 685−695)8Vallejo D, Castro-Schez J J, Glez-Morcillo C, Albusac J. Multi-agent architecture for information retrieval and intelligent mon-itoring by UAVs in known environments affected by cata-strophes. Engineering Applications of Artificial Intelligence ,2020, 87: Article No. 1032439Xie Hai-Jun, Liang Zhan-Min, Wang Jian. Design and imple-mentation of control system of UAV formation performance.Electronic Design Engineering , 2021, 29(17): 75−79(谢海军, 梁湛民, 王健. 无人机编队表演控制系统设计与实现. 电子设计工程, 2021, 29(17): 75−79)10Yang Ming-Yue, Shou Ying-Xin, Tang Yong, Liu Chang, Xu Bin. Multi-Quadrotor UAVs formation maintaining and colli-11sion avoidance control. Acta Aeronautica et Astronautica Sinica ,2022, 43: 1−11(杨明月, 寿莹鑫, 唐勇, 刘畅, 许斌. 多四旋翼无人机编队保持与避碰控制. 航空学报, 2022, 43: 1−11)Mao X, Zhang H, Wang Y. Flocking of quad-rotor UAVs with fuzzy control. ISA Transactions , 2018, 74: 185−19312Liu Ming-Wei, Gao Bing-Bing, Wang Peng-Fei, Liu Ya-Nan, Li Yi-Meng, Li Pei-Qi. Research on UAV formation obstacle avoid-ance flight based on neural network adaptive PID control. Un-manned Systems Technology , 2022, 5(2): 22−32(刘明威, 高兵兵, 王鹏飞, 刘亚南, 李怡萌, 李沛琦. 基于神经网络自适应PID 的无人机编队避障飞行控制研究. 无人系统技术,2022, 5(2): 22−32)13Li Xi-Kang, Xu Jing, Niu Yu-Gang. Memory proportional-integ-ral-retarded output sliding mode controller design. Control The-ory & Applications , 2022, 3: 1−9(李习康, 许璟, 牛玉刚. 带记忆比例−积分−时滞输出滑模控制器设计. 控制理论与应用, 2022, 3: 1−9)14Tian Bo-Lin, Li Pin-Pin, Lu Han-Chen, Zong Qun. Trajectory and attitude coordinated control of multiple unmanned aerial vehicles in complex environments. Acta Aeronautica et Astro-nautica Sinica , 2020, 41: 36−43(田栢苓, 李品品, 鲁瀚辰, 宗群. 复杂环境下多无人机轨迹姿态协同控制. 航空学报, 2020, 41: 36−43)15Yu Z Q, Liu Z X, Zhang Y M, Qu Y H, Su C Y. Decentralized fault-tolerant cooperative control of multiple UAVs with pre-scribed attitude synchronization tracking performance under dir-ected communication topology. Frontiers of Information Techno-logy & Electronic Engineering , 2019, 20(5): 685−70116Xi Wen-Long, Tang Wen-Xiu, Xu Li-Shang, Liu Fang-Yue. Pos-ition control of DC-motor based on one-order low pass filter backstepping sliding mode method. Chongqing University of Posts and Telecommunications , 2017, 29(4): 550−556(奚文龙, 唐文秀, 许李尚, 刘方悦. 基于一阶低通滤波器滑模反步法的直流电机位置控制. 重庆邮电大学学报 (自然科学版), 2017,29(4): 550−556)17Liu Z, Lou X, Jia J. Event-triggered dynamic output-feedback control for a class of Lipschitz nonlinear systems. Frontiers of Information Technology & Electronic Engineering , 2022, 23(11):1684−169918Chen Zai-Fa, Liu Yan-Cheng. Control of permanent magnet synchronous motor based on super spiral sliding model variable structure. Motor and Control Applications , 2017, 44(6): 19−23(陈再发, 刘彦呈. 基于超螺旋滑模变结构永磁同步电机的控制. 电机与控制应用, 2017, 44(6): 19−23)19Ren Yan, Wang Yi-Min, Niu Zhi-Qiang, Xiao Yong-Jian. Ap-plication of high-order terminal sliding mode control in stable platform. Control Engineering , 2021, 28(3): 553−558(任彦, 王义敏, 牛志强, 肖永健. 高阶终端滑模控制在稳定平台中的应用. 控制工程, 2021, 28(3): 553−558)20Derafa L, Benallegue A, Fridman L. Super twisting control al-gorithm for the attitude tracking of a four rotors UAV. Journal of the Franklin Institute , 2012, 349(2): 685−69921Naidu D. Singular perturbations and time scales in control the-ory and applications: An overview. Dynamics of Continuous Dis-crete and Impulsive Systems Series B , 2002, 9: 233−27822Li F, Zheng W X, Xu S Y, Yuan D M. A novel ε-dependent Lyapunov function and its application to singularly perturbed systems. Automatica , 2021, 133: Article No. 10974923He Shou-Yuan. Properties and judgment methods of positive definite matrix. Journal of Mathematical and Chemical Prob-lem Solving , 2020, 24: 18−19(何守元. 正定矩阵的性质及判定方法. 数理化解题研究, 2020, 24:18−19)24Malamud S M. A converse to the Jensen inequality, its matrix extensions and inequalities for minors and eigenvalues. Linear Algebra and Its Applications , 2001, 22(1): 19−4125Shtessel Y B, Moreno J A, Plestan F. Super-twisting adaptive26表 1 四旋翼无人机姿态角系统性能指标Table 1 Performance index of a quadrotor 'sattitude system平均收敛时间(s)平均稳态误差(rad)STW 滑模算法 2.587 1.76×10−7改进型STW 滑模算法 1.947 3.56×10−7文献[30]中的算法10.8704.24×10−68 期蔡运颂等: 基于自适应多尺度超螺旋算法的无人机集群姿态同步控制1665sliding mode control: A Lyapunov design. In: Proceedings of the49th Conference on Decision and Control. Petersburg, Russia:IEEE, 2010. 5109−5113Wang G L, Li Z Q, Miao X, Zhang Q L, Yang C Y. Fault detec-tion of discrete-time delay Markovian jump systems with delay term modes partially available. Journal of the Franklin Institute ,2019, 356(5): 3045−307127Hu Xiao-Li, Qiao Long-Kun. Improvement of Cauchy 's inequal-ity and its application. Journal of Jianghan University , 2021,49(6): 29−33(胡晓莉, 乔龙坤. 柯西不等式的改进及其应用. 江汉大学学报,2021, 49(6): 29−33)28Munoz F, Estrada M B, González-Hernández I, Salazar S, Loz-ano R. Super twisting vs modified super twisting algorithm for altitude control of an unmanned aircraft system. In: Proceed-ings of the 12th International Conference on Electrical Engineer-ing, Computing Science and Automatic Control. Tu Delft, Neth-erlands: IEEE, 2015. 1−629Jin Wan-Li, Yu Zhi-Yong, Jiang Hai-Jun. Leader-following con-sensus of second-order multi-agent systems via event-triggered impulsive control. Journal of Lanzhou University of Technology ,2022, 48(5): 153−160(金琬丽, 于志永, 蒋海军. 事件触发脉冲控制下二阶多智能体系统的领导跟随一致性. 兰州理工大学学报, 2022, 48(5): 153−160)30蔡运颂 华东理工大学信息科学与工程学院硕士研究生. 主要研究方向为滑模控制, 多智能体和无人机控制.E-mail: ********************(CAI Yun-Song Master student at the College of Information Scienceand Engineering, East China University of Science and Technology. His research interest covers sliding modecontrol, multi-agent, and UAV control .)许 璟 华东理工大学信息科学与工程学院副教授. 主要研究方向为高阶滑模观测与控制, 无人机系统建模与控制, 智能优化算法和人工智能技术.本文通信作者.E-mail: ****************.cn(XU Jing Associate professor atthe College of Information Science and Engineering,East China University of Science and Technology. Her research interest covers high-order sliding mode obser-vation and control, UAV system modeling and control,intelligent arithmetic optimization, and artificial intel-ligence technology. Corresponding author of this paper .)牛玉刚 华东理工大学信息科学与工程学院教授. 主要研究方向为随机控制系统, 滑模控制, 无线传感网络和微电网能量管理.E-mail: *****************.cn(NIU Yu-Gang Professor at the College of Information Science andEngineering, East China University of Science and Technology. His research interest covers stochastic control system, sliding mode control, wireless sensor network, and microgrid energy management .)1666自 动 化 学 报49 卷Copyright ©博看网. All Rights Reserved.。
光纤激光器
• Simulation of a nanosecond fiber laser
• Introducing a gain-guided index-antiguided fiber laser
form of a short pulse.
• The pulse duration can be shorter than the pump pulse and also shorter than the
upper-state lifetime.
• A phenomenon of spiking, where the pump power is applied for a short enough
time to generate only a single short spike of laser output.
Gain switched fiber laser based on
relaxation oscillation extraction
First pulse is generated from extracting a
so called relaxation oscillations.
These are usually damped,
eventually leading back to the
steady state.
Relaxation oscillations
Relaxation oscillation happens when
1.
2.
3.
4.
Large core area
Iliadis Modeling
1 Exactfinite-size scaling with corrections in the two-dimensional Isingmodel with special boundary conditionsW.Janke a and R.Kenna ba Institut f¨u r Theoretische Physik,Universit¨a t Leipzig,Augustusplatz10/11,04109Leipzig,Germanyb School of Mathematics,Trinity College Dublin,IrelandThe two-dimensional Ising model with Brascamp-Kunz boundary conditions has a partition function more amenable to analysis than its counterpart on a torus.This fact is exploited to exactly determine the fullfinite-size scaling behaviour of the Fisher zeroes of the model.Moreover,exact results are also determined for the scaling of the specific heat at criticality,for the specific-heat peak and for the pseudocritical points.All corrections to scaling are found to be analytic and the shift exponentλdoes not coincide with the inverse of the correlation length exponent1/ν.1.INTRODUCTIONFinite-size scaling(FSS)is a well established technique for the extraction of critical exponents fromfinite volume analyses[1].Such exponents characterise critical phenomena at a second-order phase transition.The simplest model exhibiting such a transition is the Ising model in two dimen-sions,which,despite a long history and extensive study,still offers new results and insights.Here, we study the model under the special boundary conditions of Brascamp and Kunz[2]to extract new information and to help resolve some hith-erto puzzling features of FSS.Let C L(β)be the specific heat at inverse tem-peratureβfor a system of linear extent L.FSS of the specific heat is characterized by the location of its peak,βL,its height C L(βL)and its value at the infinite-volume critical point C L(βc).The peak position,βL,is a pseudocritical point which typically approachesβc as L→∞as|βL−βc|∼L−λ,(1) whereλis the shift exponent.In two dimensions, the Ising specific heat scales as ln L.Of further interest is the FSS of the complex Fisher zeroes of the partition function[3].The leading behaviour of the imaginary part of a Fisher zero is[4]Im z j(L)∼L−1/ν,(2)where z stands generically for an appropriate function of temperature,the subscript j labels the zeroes,andνis the correlation length critical exponent.The real part of the lowest zero may be viewed as another effective critical or pseudo-critical point,scaling as|Re z1(L)−z c|∼L−λzero,(3) where z=z c atβ=βually the shift expo-nents,λandλzero,coincide with1/ν,but this is not a consequence of FSS and is not always true. The following results have been obtained for FSS in the two-dimensional Ising model.Exact Analytical Results:For toroidal lat-tices the specific-heat FSS has been determined exactly to order L−3at the infinite-volume crit-ical point in[5–7].Only integer powers of L−1 occur,with no logarithmic modifications(except for the leading term),i.e.,C L(βc)=C00ln L+C0+∞k=1C k2Numerical Results:For spherical lattices the shift exponent of the specific heat was found to be significantly away from1/ν=1,withλranging from approximately1.75to2(with the possibility of logarithmic corrections)[8].Therefore the FSS of the specific-heat pseudocritical point does not appear to match the correlation length scaling. In another study[9],FSS of Fisher zeroes for square periodic lattices yielded a value ofνwhich appeared to approach the exact value(unity)as the thermodynamic limit is approached.Small lattices appeared to yield an effective correction-to-scaling exponentω≈1.8while closer to the thermodynamic limit,these corrections tended to be analytic withω=1.A certain formal limit of conformalfield theory suggests a correction ex-ponentω=4/3[10].However,the validity of this limit has long been unclear[11]and the ques-tion of the absence of a subleading operator corre-sponding toω=4/3in the standard Ising model in two dimensions was recently addressed in depth in[12](see also[13]).In the light of these analyses,we present ex-act results which help clarify the situation.To this end,we have selected the Ising model with Brascamp-Kunz boundary conditions[2].2.FISHER ZEROESThe Brascamp-Kunz lattice has M sites in the x direction and2N sites in the y direction.The boundary conditions are periodic in the y direc-tion and the2N spins along the left and right bor-ders arefixed to...+++...and...+−+−+−..., respectively.The partition function is[2]Z∝Ni=1M j=1 1+z2−z(cosθi+cosφj) ,(5)where z=sinh2β,θi=(2i−1)π/2N andφj= jπ/(M+1).One notes that the partition function (5)is given as a double product.Determination of the Fisher zeroes of(5)is thus straightforward, as is the calculation of thermodynamic functions. For toroidal boundary conditions,on the other hand,the partition function is a sum of four such products[14].There it is non-trivial to determine the zeroes or the thermodynamic functions.The zeroes of(5)are on the unit circle in the complex-z plane(so the critical point is z c=1) [2].These are z ij=exp(iαij),whereαij=cos−1 cosθi+cosφj4 1+122−M−2σ2(1+σ2)2π 1+1M k.(10) The coefficients c k can easily be determined ex-actly and those up to c3are explicitly given in[15].So for the critical specific heat on3 a Brascamp-Kunz lattice,apart from a trivialln M/M term(which could be removed by a re-definition of M[15]),the FSS is qualitatively thesame as(but quantitatively different to)that ofthe torus topology in(4).Specific Heat near the Critical Point:Thepseudocritical point of the specific heat,z pseudoM,2N ,can be determined as the point where the deriva-tive of C M,2N(z)vanishes.This gives[15]z pseudo M,2N =1+a2ln MM2+a3ln MM3+O (ln M)2π 1+1M+d 2(ln M)2M2 ,(12)with c 0=c0and c 1=c1.Higher order terms are of the form1/M2,(ln M)2/M3,ln M/M3and 1/M3.Notice that,up to O(1/M),(12)is quanti-tatively the same as the critical specific-heat scal-ing(10).The higher order terms of(12)differ qualitatively from those in(10)in that there are logarithmic modifications of the form(ln M)k/M l (with integer k and l).Again,the values of the coefficients are given in[15].4.CONCLUSIONSFor the two-dimensional Ising model with Brascamp-Kunz boundary conditions,we have derived exact expressions for the FSS of the Fisher zeroes to all orders.We have also deter-mined the FSS of the critical specific heat,its pseudocritical point and its peak.The advantage of Brascamp-Kunz boundary conditions(over pe-riodic ones)is that the partition function is a product and meliorates determination of higher order corrections.The following are the main features we have found:All corrections to scaling are analytic(ex-cept for logarithms).The shift exponentλdoes not coincide with1/ν.The FSS of the specific-heat pseudocritical point and peak have logarith-mic corrections.Apart from the leading term, this feature is absent in the critical specific heat. REFERENCES1.M.N.Barber,in:Phase Transitions and Crit-ical Phenomena,Vol.8,eds.C.Domb and J.L.Lebowitz(Academic Press,New York, 1983),p.145.2.H.J.Brascamp and H.Kunz,J.Math.Phys.15(1974)65.3. C.N.Yang and T.D.Lee,Phys.Rev.87(1952)404;ibid.410;M.E.Fisher,in:Lectures in Theoretical Physics,Vol.VIIC,ed.W.E.Brit-tin(Gordon and Breach,New York,1968), p.1.4. C.Itzykson,R.B.Pearson,and J.B.Zuber,Nucl.Phys.B220(1983)415.5. A.E.Ferdinand and M.E.Fisher,Phys.Rev.185(1969)832.6.N.Sh.Izmailian and C.-K.Hu,Phys.Rev.E(in print)[cond-mat/0009024].7.J.Salas,J.Phys.A34(2001)1311.8.J.Gonz´a lez and M.A.Mart´ın-Delgado,hep-th/9301057;O.Diego,J.Gonz´a lez,and J.Salas,J.Phys.A27(1994)2965;Ch.Hoel-bling and ng,Phys.Rev.B54(1996) 3434.9.N.A.Alves,J.R.Drugowich de Felicio,andU.H.E.Hansmann,Int.J.Mod.Phys.C8 (1997)1063.10.J.Zinn-Justin,Quantum Field Theory andCritical Phenomena,3rd ed.(Clarendon Press,Oxford,1996),p.636.11.G.Jug and B.N.Shalaev,J.Phys.A32(1999)7249.12.P.Calabrese,M.Caselle,A.Celi,A.Pelis-setto,and E.Vicari,J.Phys.A33(2000) 8155.13.M.Caselle,M.Hasenbusch,A.Pelissetto,andE.Vicari,cond-mat/0106372.14.B.Kaufman,Phys.Rev.76(1949)1232.15.W.Janke and R.Kenna,Phys.Rev.B(inprint)[cond-mat/0103332].。
Analysis of Delay Test Effectiveness with a Multiple-Clock Scheme
Analysis of Delay Test Effectiveness with a Multiple-Clock Scheme Jing-Jia Liou,Li-C.Wang,and Kwang-Ting ChengDepartment of ECE,UC-Santa Barbarajjliou,licwang,timcheng@Jennifer Dworak,and M.Ray MercerDepartment of EE,Texas A&M Universityjdworak,mercer@Rohit Kapur,and Thomas W.WilliamsSynopsys Inc.rkapur,tww@AbstractIn conventional delay testing,two types of tests,transition tests and path delay tests,are often considered.The test clock fre-quency is usually set to a single pre-determined parameter equal to the system clock.This paper discusses the poten-tial of enhancing test effectiveness by using multiple test sets with multiple clock frequencies.The two intuitions motivating our analysis are1)multiple test sets can deliver higher test quality than a single test set,and2)for a given set of AC de-lay patterns,a carefully-selected,tighter clock would result in higher effectiveness to screen out potentially defective chips. Hence,by using multiple test sets,the overall quality of AC de-lay test can be enhanced,and by using multiple-clock schemes the cost of adding the additional pattern sets can be minimized. In this paper,we analyze the feasibility of this new delay test methodology with respect to different combinations of pattern sets and to different circuit characteristics.We discuss the pros and cons of multiple-clock schemes through analysis and ex-periments using a statistical delay evaluation and delay defect-injected framework.1IntroductionIn traditional AC delay test and validation,two sets of patterns are often applied.They are the transition fault patterns and the critical path test patterns.Conventional wisdom usually dif-ferentiates the two by the sizes of delay defects they intend to capture.On one hand,transition fault patterns are considered to be more effective for large-size delay defects which can hap-pen randomly at any site of a circuit.On the other hand,criti-cal path test patterns aim to detect small-size delay defects on a selected set of long timing paths.It is then hoped that the com-bination of these two orthogonal strategies can capture most of the delay defects and ensure the circuit performance.By definition,transition fault test generation does not rely on a timing analysis tool.For a given site,delay faults can be observed through any of the sensitizable paths.Hence,the delay defect sizes captured by transition fault tests are not guaranteed.The longer the timing of the paths used to de-tect the faults,the smaller the sizes of defects they can cap-ture.Without explicitly targeting specific paths,employing multiple-detections becomes an alternative for enhancing the quality of transition test set effectiveness[5].The quality of path delay tests,on the other hand,depends on the use of a timing analysis tool to accurately estimate the timing lengths of the paths.Assuming an ideal timing analy-sis tool is in hand,a set of critical paths can be derived.AC patterns are then produced to specifically test these paths.Test generation for target paths is not a trivial task,especially for paths whose robust detection is not possible[10].Hence,an ideal critical path set does not imply that an ideal test pattern set can be generated.In traditional timing analysis,delay models are often dis-crete and based upon nominal or worst-case timing assump-tions[3,4,6,7].Such models become increasingly inadequate because,in the deep sub-micron domain,factors affecting de-lay characteristics(such as process variations,manufacturing defects,and noise)can often be continuous in nature[1,2,8,9]. These continuous factors can be better captured and simulated using statistical models and methods[11,12,13].Hence,de-lays should be modeled with random variables instead of with discrete values.This adds tremendous complexity to path delay fault testing.With statistical delay models in mind,the focus of this paper does not lie in maximizing the test effectiveness for transition fault patterns nor in selecting the true longest paths for path de-lay testing.Instead,we explore another alternative to improve delay test quality–by applying the test patterns in phases with different clock frequencies.We assume that a set of test pat-terns is already given.Then,the question is how to apply an additional test set(s)to further enhance the test quality with minimal additional cost.It is important to note that,with the methodology introducedin this paper,applying the same patterns at different clock fre-quencies will not enhance the test quality of a given pattern set(This point will be further illustrated in Section3).Hence, to improve test quality,our strategy is the following:Given a set of patterns,a second set of patterns is added.Test qual-ity is improved by combining the two pattern sets.However, the application of tests is divided into two phases:the screen-ing phase and the confirmation phase.In the screening phase, the goal is to use thefirst set of patterns with a tighter clock scheme to screen out potentially defective chips.Then,in the confirmation phase,both sets of patterns are applied with the normal clock to confirm which chips are actually bad.The key of this strategy is the selection of the clock scheme in the screening phase in order to minimize the cost during the con-firmation phase.The goal is to achieve a test quality similar to that obtained by applying both test sets in the traditional way and yet,with a cost closer to that of applying only thefirst test set alone.Our earlier work[14]proposed using the2-phase test scheme based upon double-transition test set,and concluded that during the screening phase,using only one tighter clock might not be enough.In this paper,we discuss the potential of applying a multiple-clock scheme(where more than one tighter clocks are used during the screening phase)with different com-binations of test sets and with designs of different timing char-acteristics.We explore the advantages and limitations of using the multiple-clock scheme for enhancing delay test effective-ness.Our conclusion is that without an ideal timing analysis tool and a perfect path-delay ATPG tool,combining multiple-clock scheme and multiple-detection transition test sets will be a feasible solution in practice for improving delay test quality, especially for timing-optimized high-performance designs.This paper is organized as the following.Section2de-scribes the background and motivation for our work.Section3 presents the methodology of using a multiple-clock scheme for test quality improvements and discusses related issues.Sec-tion4illustrates our experimental setting and presents experi-mental results based on an ISCAS benchmark circuit.In sec-tion5,we analyze the feasibility of the proposed clock scheme with respect to different combinations of test sets and to differ-ent circuit characteristics.Section6concludes the paper.2Background and Motivation Several publications[15,16,17,18]have tried to use variable or multiple clock schemes to improve the detection of delay de-fects.Unfortunately,there is one serious drawback with these schemes:the false negative outcomes.It is possible that most chips considered“bad”by these schemes are actually function-ally correct and have correct timing.This is because most delay defects that increase the delays of some paths may not actually be located on the critical timing paths.This problem will re-sult in an unnecessary increase in the number of parts deemed defective and a decrease of the total yield(yield loss).There-fore,for practical use of a multiple-clock scheme,thefirst key concern is to avoid yield loss.In conventional AC delay test,the transition fault model and path delay model represent two orthogonal testing strategies.Assume that a set of100%transition patterns sensitizes thepaths P tran l1l n and a set of path delay patterns sensi-tizes P path p1p k.Let M be an ideal timing model such that given a path p,M p denotes the delay of the path.And let C be the clock.Then,by covering P tran,all the single-sited de-fects with sizes larger thanθtran C min M l1M l n are expected to be captured.Similarly,all the defects with sizes larger thanθpath C min M p1M p k will be captured if the defects fall on a path in P path.If M is a statistical model, we note that bothθtran andθpath are random variables,not sin-gle values.In general,we should haveθtranθpath.In transition fault testing,P tran does not always contain the longest path passing through a site.To improve transition test quality,it is intuitive to improve the quality of P tran by sensitiz-ing the path with longest propagation time l maxipassing through each site i.However,the cost of doing so will be similar to that of path delay testing,and the quality will once again be highly dependent on the timing model M.The quality of P path depends on the accuracy of M.In thedeep sub-micron domain,the accuracy of M is not easy to de-fine,and traditional timing analysis will often fail to provide a good delay prediction.The topological coverage of P path may also affect the quality.The topological coverage of P path con-cerns the site coverage.If P path covers every site,then any single-sited delay defect,in theory,will be captured.In this case,those defects that are not captured will not affect the cir-cuit performance.However,normally,it is not expected that P path covers every site.(In that case,P path could replace tran-To illustrate the above analysis,Figure1demonstrates com-parison results between transition patterns and path delay pat-terns.These results were obtained using the statistical timing analysis framework developed before[13].For a particular de-fect size,1000randomly-located defects are injected onto the circuit(this is ISCAS benchmark s5378).Then,we collect the statistics of unique defect detection.In thefigure,two pattern combinations are analyzed:single-detection transition patterns(tfp1)vs.path delay fault patterns (pdf),and double-detection transition patterns(tfp2)vs.thesame path delay patterns.Here,we are interested in the uniquedetection within each combination.For example,in the case of defect size15units,(if they are detected)there is no defect uniquely detected by the single-detection transition patterns. All detected defects are captured through the path delay pat-terns.For the double-detection transition patterns,still there is no unique detection.However,the unique detection from the path delay patterns drops due to a higher overlap of the detec-tion between the double-detection patterns and the path delay patterns.In general,several things can be observed from thefigure: Transition fault patterns are better for large-size defects (contribute more unique detections when the defect sizes are large).Path delay patterns are better for small-size defects,and since the patterns do not cover all sites in the circuit,for large defects its value can saturate.(Once a defect fallsbeyond the topological coverage,it has no chance of be-ing detected.)Double-detection transition patterns are better thansingle-detection transition patterns.In addition,because of the higher overlap of detection between the double-detection patterns and the path delay patterns,the uniquedetection of the path delay patterns decreases.The benefit of using double-detection transition patterns is clearly shown in thefigure.However,the major problem ofusing double-detection patterns is its high test application cost. Normally,for state-of-the-art high-performance designs such as microprocessors,applying single detection patterns wouldhave already consumed a majority of the test application re-sources(tester memory,time,etc.).It is hence very difficult to adopt another complete transition pattern set in practice.The strength of a transition pattern set lies in its complete topological coverage of the circuit.Its weakness is the ability to detect small-size defects.Multiple-detection can enhancethat ability with a relatively high cost.Is there any other alter-native so that we can keep the benefit of going for a multiple-detection strategy but without the high cost associated with it?Our answer is based upon alternating the test clock C.In-stead offixing the test clock to be the same as the system clock,we can carefully devise a test clock scheme which con-sists of multiple clocks.For example,when pattern set1is applied,instead of using C we may use a different clock Cδ. By doing so,any single-sited defect with a size greater thanθalt Cδmin M l11M l1n will be screened out.Be-causeθaltθtran,smaller defects that were not detectable be-fore can be captured now.Hence,we have enhanced the ability to screen out“defective”chips.Notice that by applying tests with a tighter clock,a good chip passing the new clock Cδwould still be classified as a good chip if the same set of pat-terns is applied with the original clock C.However,the reverse is not true:a failing chip may still pass the same test set with C.In this case,we may have a yield loss.In the next section,we will discuss how a multiple-clock scheme should be devised to avoid the problem of yield loss and at the same time improve quality with minimal cost.Then,throughout the rest of the paper,we will focus our discussion on the following two issues:What combination(s)of multiple test sets will be mosteffective to be applied with a multiple-clock scheme?What circuit characteristics are most suitable for using amultiple-clock scheme?We will analyze these issues based upon transition fault testand path delay fault test strategies,using both un-optimizedand optimized circuits,and under a statistical delay evaluation and defect-injected simulation framework.3Multiple-Clock SchemeFor simplicity of discussion,we assume two100%sets of tran-sition fault patterns A and B are given.Let A cover pathsT1l11l1n,and let B cover paths T2l21l2n for sites1n,respectively.Our goal is tofind a clock scheme such that the cost is minimized relative to the cost of applying onlyone set,and the quality is maximized relative to the quality ofapplying both sets.Our methodology employs a2-phase test application pro-cess:a screening phase and a confirmation phase.During thescreening phase,tighter clocks are used with the application of test set A to quickly screen out the potentially bad chips.Then, during the confirmation phase,the normal clock is used with the test set A B to avoid yield loss.We note that,to avoid yield loss,it is necessary to applyboth A and B with the normal clock during the confirmationphase in order tofilter in good chips that were considered to be defective during the screening phase.Therefore,with this methodology,it is important to note that alternating the clock by itself will not improve the quality beyond that obtained by applying both A and B with the normal clock.Hence,A B combined with the normal clock will represent an upper bound (in terms of test quality)for using a multiple-clock scheme.In other words,the total number of defective chips captured in the new scheme cannot be greater than that captured by the original A B combined with the normal clock.Assume that the tighter clock is Cδ.The following dis-cusses the concepts of defect miss and yield loss in more detail. Defect Miss A chip passing the screening phase may be defec-tive and still could be captured if B were applied with the normal clock.However,it is guaranteed that the passed chip will still pass if A is applied with the normal clock.Based upon these points,it is possible that some defectsare missed under the new test application scheme butcould be captured by the original test method where A B is applied with the normal clock.How likely is this to happen?Consider a given site i.A true defect(which can affect circuit performance)on iwill be missed if its size d falls into the range C M l2iCδM l1i.This implies M l1iδC d M l2i(see Figure2).With a reasonably largeδ,it is unlikelythat M l2i M l1iδ.Even when this is happening, the chance of a defect occurrence with a size that fallsexactly into the range is small.Hence,the problem of a defect miss is not a serious concern.One may think that to avoid defect miss,we should set δas large as possible.However,a large δ(a tighter clock)will increase the potential yield loss as explained below.1+1++Figure 2:Illustration of Defect Miss and Yield LossPotential Yield Loss Yield loss is defined as the percentage ofchips that are considered to be defective by the tests but are actually good chips.In our scheme,since we have al-ready used A B and the normal clock in the confirmation phase to prevent yield loss from happening,the key here is how likely will potential yield loss happen at the end of screening phase?In other words,if many chips are deter-mined to be bad in the screening phase,then these chips will be tested by A B again.Hence,the higher such a potential yield loss is,the higher the cost of the proposed testing scheme will be.Similarly to the defect miss situation,a defect at site i will result in potential yield loss (at the end of the screen-ing phase)if its size d falls into the range C δM l 1i C max M l 1i M l 2i (see Figure 2).This implies thatmax M l 1i M l 2i C d δM l 1i .Hence,to min-imize potential yield loss during the screening phase (and hence the overall test application cost),δshould be as small as possible.From the figure,we note that for a fixed δand a fixed site i ,a defect miss and potential yield loss cannot happen at the same time.The Selection of δThe above analysis suggests two contradictory objectives for the selection of δ.On one hand,δshould be large to avoid a defect miss.On the other hand,δshould be small to mini-mize the test cost resulting from the effort of avoiding yield loss during the confirmation phase.By combining the two ob-jectives,it seems that the most logical solution is to set a differ-ent δi M l 1i M l 2i for each site i .This is obviously not an easy thing to accomplish because we may need to supply so many different test clocks during testing,and even though that is possible with a high cost,ensuring that the timing model M accurately reflects the reality is also difficult.3.1Clock Scheme in the Screening PhaseThe compromise solution is to employ k multiple clocks in the screening phase,where k is a small number.Our strategy is to first partition a given test set (A )in the screening phase into k subsets.Then,for each subset,a different clock is determined.Test Set Partitioning Given a set of patterns,we use thestatistical timing analysis tool M [13]to characterize the delay distribution of the longest sensitizable path passing through each site (again denote these results asM l 11M l 1n ).Next we compute the mean value in each distribution (assume that the order is m l 11m l 1n).Then,the partition is based upon the difference D m l 1n m l 11.The first partition is to divide 0D into two ranges D 10εD and D 2εD D where εdenotes the ratio of partition and 0ε1(normally,05εbecause short paths are less important than the long paths).All paths whose mean timing values fall into the first range will be grouped into the first set S 1.The rest will be grouped into the second set S 2.Then,the same idea can be applied recursively to partition S 2fur-ther,and so on (see Figure 3).It is interesting to note that in our experiments,we discovered that the best results are often obtained by setting:ε11provide that coverage.The key reason is that a tighter clock will not help to capture a defect if the defect falls beyond the topological coverage of a pattern set.Withouta complete topological coverage,the number of defectmisses could be high.The multiple-clock scheme alone will not help to improve test quality.The improvement of quality actually comes from the addition of the second pattern set.The usage of the multiple-clock scheme is to greatly reduce the cost associated with the inclusion of the second pattern set.In essence,the goal of the screening phase is to quickly identify potentially defective chips.Then,in the confir-mation phase,a higher quality test set is applied to double check each potentially bad chip more carefully.The cost saving relies on the fact that the majority of the chips are good ones and hence,only a few chips would go into the confirmation phase.3.3Cost EvaluationGiven the two pattern sets,A and B,intuitively we may define the cost as AρA B whereρrepresents the probability of a chip classified as potentially defective during the screening phase.As discussed before,this probabilityρdepends on the clock(s)selected during the screening phase.In the next sec-tion,we will present ourflow to evaluateρunder the statisti-cal timing analysis framework.However,the true cost is not entirely reflected in the simple equation AρA B.This is because the calculation of cost should also take defect size distribution into account.Without loss of generality,we as-sume defect distribution is a discrete functionΓ.Then,for each defect size s,Γs gives the conditional probability that if a defect does occur,its size is s.Then the true cost should be re-formulated as:Cost A∑sΓsρs A B(1) where s is the defect size andρs is the probability of a chip being classified as a defective chip during the screening phase due to a defect with size s.In essence,equation(1)calculates the cost by averaging across the defect distribution curve.If the defect distribution is uniform,then the equation will be reduced to“AρA B,”whereρ∑sρs.3.4Cost Vs.Quality GainWe note that in the proposed multiple-clock methodology,the quality of the overall test scheme is determined by the quality of“A B.”The cost depends onρand the size of“A B.”Suppose the size of B is relatively small with respect to A,and the quality of B is very high.Then,using the multiple-clock method will have little gain in this case because A A B. In other words,using the multiple-clock scheme will not be better than just applying A B together with a normal clock.As mentioned before,A has to be a transition fault pattern set to ensure a complete topological coverage.For B,if a small-size high-quality path delay pattern set is available,then the multiple-clock scheme will be of little usefulness.However,we emphasize that the quality of a path delay pattern set de-pends on the timing tool and the test generation tool,whichboth are far from being perfect in reality.Moreover,the sizeof the path delay pattern set needed to achieve a desired highquality also depends on the circuit’s timing characteristics as well.Hence,in section5we will discuss how the circuit’s tim-ing characteristics may affect the effectiveness of the proposedmultiple-clock scheme.3.5Cost ReductionThe cost can be further reduced if we allow an extremely smallyield loss tolerance level.For example,we can select an upperbound t to restrict the target defect size during the confirmation phase such that for any large-size defect s,s t,andΓsεwhereεis a very small number such as104.Then,all shortpaths whose timing lengths less than C t(with high proba-bilities)in A can be removed during the confirmation phase. The intuition behind this is that these short paths will be of lit-tle help in deciding whether a potential defective chip after the screening phase is actually a good chip or not.Accordingly, we can remove all patterns that cover only those removed short paths.In the next section,we will demonstrate that ifΓis an exponential distribution,a large number of paths in A can be removed in the confirmation phase with very little increase(al-most zero)of the yield loss probability.As a result,the cost of applying our proposed test scheme can be dramatically re-duced,and the quality can be maintained.4Experiments4.1Evaluation FrameworkCoverageanalysisofSonthegivencircuitinstanceFigure4:Flow Chart for Statistical Evaluation of S We implemented a statistical delay evaluation framework to estimate the defect coverage based on simulation of random defect injection.Figure4illustrates the complete procedure of our evaluation scheme for a particular clock and pattern set. Note that the pattern set is characterized using the set of paths S sensitized by the patterns.In each Monte Carlo sampling run,first a circuit instance with cell/interconnect delays is gener-ated according to the delay distributions characterized through Monte Carlo SPICE.This instance will then be evaluated by“statistical analysis of S”.The “statistical analysis of S”is used to check if there is any path in S (on the given instance)longer than the testing clock C .If there is,then this instance is said to be faulty and covered by S (Captured ).At the end,our scheme will calculate the capture probability for S .For the multiple-clock scheme,the evaluation process is similar.Those circuit instances classified as Captured during the screening phase will further be tested in the confirmation phase.Then,at the end,a Captured probability number will be produced.This statistical delay evaluation framework requires pre-characterization of cells,i.e.,building libraries of pin-pin cell delays and output transition times (as random variables).In our experiments,we utilize a Monte-Carlo-based SPICE (ELDO)[19]to extract the statistical delays of cells for a 0.25µm,2.5V CMOS technology.The input transition time and output loading of the cells are used as indices for build-ing/accessing these libraries.4.2Initial ResultsIn the initial experiments,three test sets are considered:two single transition fault test sets,and a path delay fault set.For the path delay fault set,we will select statistically long paths [13]which are functionally sensitizable [10].Without a perfect path delay test generator,we assume one test for each selected path.In a sense,this represents the optimal situation (there ex-ists a test for every selected path and an ATPG tool can always produce the test).In the following,we explain results from circuit s5378.Fig-ures 5and 6present test quality levels that resulted from the multiple-clock scheme.In the figures,we demonstrate,for each method,its ability to capture defects of various sizes.We note that since our objective here is to compare the rela-tive strength among different methods,the absolute values oftern Sets Using Different Clock SchemesIn Figure 5we combine th two transition fault pattern sets.In Figure 6we combine the first transition fault pattern set with the path delayset where the (statistically)longest 40paths are selected.For the multiple-clock scheme in the screening phase,Path Delay Pattern Sets Using Different Clock Schemes the partition number k is set to 3,and the ratio of partition εis set to 114.3Cost ComparisonTF2TF1-20.460.72cost(%) 1.006 2.006 reduced cost t150 1.006 2.0061.000 1.027reduced cost t50 1.006 2.006 TF1:1st transition set with normal clockTF2:2nd transition set with normal clock(TF1-2):two sets combined with the multiple clock schemeTF1-2:two sets combined with normal clock(TF1/PDF)capture prob.0.82cost(%) 1.1420.033 1.0330.033 1.0330.033 1.033PDF:path delay fault set with normal clock(TF1/PDF):1st transition set path delay set with multiple clocksTF1/PDF:1st transition set path delay set with normal clock Table1:Test cost and capture probability comparison for dif-ferent delay fault sets and clock schemes for s5378.Table1demonstrates the capture probabilities and cost com-parison results for different testing schemes.The costs are cal-culated by the formula in section3.3and based on the assump-tion of a defect size distribution:λeλs where s is the defect size andλis a constant(we useλ01in the experiments). This exponential distribution for defect size(given that defects occur)has been studied in many publications[20,21]and is a practical assumption to be used.Note that it is also possible to adopt other distributions.However,using other distributions in general does not invalidate the trends observed in our analysis.In the table,we also consider the cost reduction approach discussed in section3.5.The t value in each case is shown in the table.Note that by assuming the given defect distribu-tion,the probability of yield loss is0.000676for the worst case shown in the table(where t50units).We note that by def-inition,the yield loss is zero without involving the cost reduc-tion.Hence,this demonstrates that the proposed cost reduction scheme has a very small(but non-zero)probability of incurring a yield loss.We note that the costs presented in the table are normalized with respect to the cost of TF1.Data in this table confirm our earlier observations:The quality that resulted from the multiple-clock scheme “(TF1-2)”is close to the quality that resulted from two sets combined“TF1-2.”When using two transition test sets,the cost without re-duction can be high(127.4%).However,with cost reduc-tion,the cost can be reduced to almost the same as TF1.Notice that in case of cost reduction(meaningful only for (TF1-2)),by linearly decreasing the t value,the cost can drop in terms of an order of magnitude.In this analysis,a high-quality small-size path delay test set could be obtained.Hence,notice that in this case,themultiple-clock scheme does not provide much benefit,as expected.In reality,obtaining a high-quality small-size path delay set is very hard.Hence,multiple transition test sets combined with the multiple-clock scheme provides a much easier and practi-cal alternative to enhance test quality.However,in the table, we see that the quality of combining two transition test sets re-mains far from that achievable by an ideal path delay test set. Results from this table motivate us to further the study of the effectiveness of the multiple-clock scheme.The key question now is:Under what condition can multiple transition test sets combined with the multiple-clock scheme approach the cost and effectiveness of a high-quality path delay test set?Our intuition is that the effectiveness of using the multiple-clock scheme depends upon the circuit’s timing characteristics.5In-Depth AnalysisIn this section,we investigate how the circuit’s timing char-acteristics may affect the effectiveness of the multiple-clock scheme.The goal is to identify the most suitable circuit styles to be applied with the proposed scheme.For example,a highly timing-optimized circuit with a shallow logic depth may be-have quite differently under the proposed scheme from a cir-cuit that has not been optimized for timing.To demonstrate the difference,two additional experiments were devised.Thefirst experiment demonstrates the effect of the timing optimization process on the effectiveness of using the multiple-clock scheme.Given the benchmark circuit s5378,we measure the quality results based upon different optimized versions of the design.In the second experiment,a highly optimized in-dustrial design is used to emphasize the timing optimization effect.5.1The Impact of Timing Optimization We generate three different versions of circuit s5378by apply-ing different timing optimization constraints in Synopsys De-sign Compiler[22].These three versions are denoted as“v1,”“v2,”and“v3,”where v3is most optimized,and v1is the orig-inal version without timing optimization.Our statistical timing analysis tool is used to characterize the path delay profiles for the three circuits.These profiles are shown in Figure8.Note that only functionally sensitizable paths are counted in the profiles.In the optimization process, we try to improve not only the maximum delays of the circuit but also the span of the delay spectrum.Both parameters are smaller in each succeeding version(v1v2v3).Again,we consider two transition fault test sets combined with the multiple-clock scheme.In addition,we produce three more sets,PDF10,PDF50,PDF100which correspond to the(statistically)longest10,50,and100path sets,respec-tively.These selected paths are functionally sensitizable,and we again assume an ideal test generator is available to produce one test for each path.For the three versions of the circuit,the system clocks should be different as well to ensure a fair comparison.This is because v3can operate at a higher speed than v1.If the same。
Frank copula R程序语句
### R code from vignette source 'Frank-Rmpfr.Rnw'###################################################### code chunk number 1: preliminaries###################################################op.orig <-options(width = 75,SweaveHooks= list(fig=function() par(mar=c(5.1, 4.1, 1.1, 2.1))),useFancyQuotes = FALSE,## for JSS, but otherwise MM does not like it:## prompt="R> ",continue=" ")# 2 (or 3) blanks: use same length as 'prompt' copDDir <- system.file('doc', package='copula')Sys.setenv(LANGUAGE = "en")if(.Platform$OS.type != "windows")Sys.setlocale("LC_MESSAGES","C")###################################################### code chunk number 2: nacopula-dDiagA (eval = FALSE)##################################################### copula:::dDiagA###################################################### code chunk number 3: nacopula-dDiagA-show###################################################writeLines(head(capture.output(print(copula:::dDiagA)), -1))###################################################### code chunk number 4: my-dDiagA###################################################dDiagA <- function(u, th, d, iPsi, absdPsi, absdiPsi, log = FALSE) { stopifnot(is.finite(th), d > 0, is.function(iPsi),is.function(absdPsi), is.function(absdiPsi))if(log) {log(d) + absdPsi(d*iPsi(u,th), th, log = TRUE) +absdiPsi(u, th, log = TRUE)} else {d* absdPsi(d*iPsi(u,th), th) * absdiPsi(u,th)}}###################################################### code chunk number 5: def-orig-psi-func###################################################iPsi.0 <- function(u,theta) -log( (exp(-theta*u)-1) / (exp(-theta)-1) ) iPsi.1 <- function(u,theta) -log(expm1(-u*theta) / expm1(-theta)) absdiPsi.1 <- function(u, theta, log = FALSE)if(log) log(theta)-log(expm1(u*theta)) else theta/expm1(u*theta)###################################################### code chunk number 6: def-orig-absdPsi###################################################require("copula")# for polylog()absdPsi.1 <- function(t, theta, log=FALSE) {p <- -expm1(-theta)Li. <- polylog(log(p) - t, s = 0, log=log,method="negI-s-Eulerian", is.log.z=TRUE)if(log) Li. - log(theta) else Li. / theta}###################################################### code chunk number 7: simpler-absdPsi###################################################absdPsi.2 <- function(t, theta, log=FALSE) {w <- log(-expm1(-theta)) - tLi. <- if(log) w - log(-expm1(w)) else -exp(w)/expm1(w)if(log) Li. - log(theta) else Li. / theta}###################################################### code chunk number 8: dDiag-probl-big-theta###################################################curve(dDiagA(x, th = 38, d = 2, iPsi = iPsi.0,absdPsi=absdPsi.1, absdiPsi=absdiPsi.1, log = TRUE),ylab = "dDiagA(x, theta= 38, *, log=TRUE)",0, 1, col = 4, n = 1000)## and using the slightly better iPsi.1 does not help here:curve(dDiagA(x, th = 38, d = 2, iPsi = iPsi.1,absdPsi=absdPsi.2, absdiPsi=absdiPsi.1, log = TRUE),add = TRUE, col = 2, n=1000)legend("bottom", c("iPsi.0()","iPsi.1()"),col=c(4,2), lty=1, bty="n")###################################################### code chunk number 9: dDiag-ok-big-theta###################################################iPsi.2 <- function(u,theta) -log1p((exp(-u*theta)-exp(-theta)) / expm1(-theta))curve(dDiagA(x, th = 38, d = 2, iPsi = iPsi.2,absdPsi=absdPsi.2, absdiPsi=absdiPsi.1, log = TRUE),ylab = "dDiagA(x, theta= 38, *, log=TRUE)",0, 1, col = 4, n = 1000)## previouslycurve(dDiagA(x, th = 38, d = 2, iPsi = iPsi.1,absdPsi=absdPsi.2, absdiPsi=absdiPsi.1, log = TRUE),add = TRUE, col = "darkgray", lwd=2, lty=3, n=1000)###################################################### code chunk number 10: ex-U-data###################################################d <- 5(theta <- copFrank@tauInv(tau = 0.75))cop <- onacopulaL("Frank", list(theta, 1:d))set.seed(1); for(l in 1:4) U <- rnacopula(n = 100, cop)U. <- sort(apply(U, 1, max)) # build the max###################################################### code chunk number 11: negLogL###################################################mlogL <- function(theta)-sum(dDiagA(U., theta, d=d, iPsi = iPsi.2,absdPsi=absdPsi.2, absdiPsi=absdiPsi.1,log = TRUE))###################################################### code chunk number 12: llog-theta-plot###################################################p.mlogL <- function(th, mlogL, col= "red2", lwd = 1, lty = 1,add= FALSE) {stopifnot(is.numeric(th), is.function(mlogL))nll <- vapply(th, mlogL, 0.)if(add) lines(nll ~ th, col=col, lwd=lwd, lty=lty)else plot(nll ~ th, xlab=expression(theta),ylab = expression(- logLik(theta, .)),type = "l", col=col, lwd=lwd, lty=lty)invisible(nll) # return invisibly}thet <- seq(11, 99, by = 1/4)p.mlogL(thet, mlogL)require("Rmpfr")## compute the same with *high* accuracy ...## using three different precisions:MPrecBits <- c(160, 128, 96)mkNm <- function(bits) sprintf("%03d.bits", bits)## As it takes a while, cache the result:fnam <- sprintf("mlogL_mpfr_%s.rda", ()[["machine"]])if (!file.exists(fn <- file.path(copDDir,fnam))) {print(system.time(nllMP <- lapply(MPrecBits, function(pBit) {nlM <- thM <- mpfr(thet, precBits = pBit)## (vapply() does not work for "Rmpfr":)for(i in seq_along(thet)) nlM[i] <- mlogL(thM[i])nlM}))) ## 91.226 0.013 91.506 [nb-mm icore 5]names(nllMP) <- mkNm(MPrecBits)copSrcDDir <- if(Sys.getenv("USER") == "maechler")'~/R/Pkgs/copula/inst/doc' else ""if(file.exists(copSrcDDir))# <<- only for certain users; not on CRAN etcsave(nllMP, file = file.path(copSrcDDir, fnam))} else load(fn)colB <- c("blue3","violetred4","tan3")ltyB <- c(5:3)lwdB <- c(2,2,2)for(i in seq_along(nllMP)) {lines(thet, as.numeric(nllMP[[i]]),col=colB[i], lty = ltyB[i], lwd = lwdB[i])}leg <- c("double prec.", sprintf("mpfr(*, precBits = %d)", MPrecBits))legend("top", leg,col= c("red3",colB), lty=c(1, ltyB), lwd=c(1,lwdB), bty="n")###################################################### code chunk number 13: empty-prompt###################################################op <- options(prompt = " ")###################################################### code chunk number 14: mlogL_2terms (eval = FALSE)##################################################### absdPsi(d*iPsi(u,th), th, log = TRUE) +## absdiPsi(u, th, log = TRUE)###################################################### code chunk number 15: 3f (eval = FALSE)##################################################### iPsi = iPsi.2; absdPsi = absdPsi.2; absdiPsi = absdiPsi.1###################################################### code chunk number 16: reset-prompt###################################################options(op)###################################################### code chunk number 17: check-iPsi###################################################stopifnot(all.equal(iPsi.2(U., 50 ),iPsi.2(U., mpfr(50, 96))),all.equal(iPsi.0(U., mpfr(50, 200)),pI.U <- iPsi.2(U., mpfr(50, 200)), tol=1e-50) )###################################################### code chunk number 18: plot-absdPsi-2###################################################psD.n <- absdPsi.2(as.numeric(pI.U), 40)psD.M <- as.numeric(absdPsi.2(pI.U, mpfr(40, 200)))matplot(U., cbind(psD.n, psD.M), type="l", log="y")legend("top", c("double prec.", "mpfr(*, precBits = 200)"), col= 1:2, lty=1:2, bty="n")###################################################### code chunk number 19: plot-absdPsi-zero###################################################u0 <- 2^-(100:1)psD.n <- absdPsi.2(u0, 40)psD.M <- as.numeric(absdPsi.2(u0, mpfr(40, 200)))matplot(u0, cbind(psD.n, psD.M), type="l", log="xy")legend("top", c("double prec.", "mpfr(*, precBits = 200)"), col= 1:2, lty=1:2, bty="n")###################################################### code chunk number 20: log1mexp###################################################log1mexp <- function(a) # accurately compute log(1-exp(-a)) {stopifnot(a >= 0)r <- atst <- a <= log(2)r[ tst] <- log(-expm1(-a[ tst]))r[!tst] <- log1p(-exp(-a[!tst]))r}###################################################### code chunk number 21: absdPsi.3################################################### absdPsi.3 <- function(t, theta, log=FALSE) {w <- log1mexp(theta) - tLi. <- if(log) w - log1mexp(-w) else -exp(w)/expm1(w)if(log) Li. - log(theta) else Li. / theta}###################################################### code chunk number 22: nlogL-2-plot###################################################p.mlogL(th = seq(11, 99, by = 1/4),mlogL = (mlogL2 <- function(theta)-sum(dDiagA(U., theta, d=d, iPsi = iPsi.2,absdPsi = absdPsi.3, absdiPsi=absdiPsi.1, log = TRUE))), lwd = 2)###################################################### code chunk number 23: llog-theta-plot-2###################################################thet <- 9:1000nll <- p.mlogL(thet, mlogL = mlogL2, lwd=2)(th0 <- thet[i0 <- max(which(is.finite(nll)))])abline(v = th0, col="red2", lty="15", lwd=2)###################################################### code chunk number 24: dDiag-large-thet###################################################dDiagA(0.999, 715, d = d, iPsi = iPsi.2, absdPsi = absdPsi.3,absdiPsi = absdiPsi.1, log = TRUE)###################################################### code chunk number 25: psiID1-large-thet###################################################absdiPsi.1(0.999, th = 715, log=TRUE)###################################################### code chunk number 26: def-absdiPsi-2###################################################absdiPsi.2 <- function(u, theta, log = FALSE)if(log) log(theta)- {y <- u*theta; y + log1mexp(y)} elsetheta/expm1(u*theta)###################################################### code chunk number 27: llog-theta-plot-3###################################################plot(nll ~ thet, xlab=expression(theta),ylab = expression(- logLik(theta, .)),type = "l", col="red2", lwd=2)abline(v = th0, col="red2", lty="15", lwd=2)nll3 <- p.mlogL(thet, mlogL = function(theta)-sum(dDiagA(U., theta, d=d, iPsi= iPsi.2, absdPsi= absdPsi.3,absdiPsi = absdiPsi.2, log = TRUE)), col = "blue2", lwd=3, lty=2, add = TRUE)nll3[thet == 800]###################################################### code chunk number 28: true-iPsi-small###################################################pI <- iPsi.2(u=0.999, th= mpfr(800, 200))cat(sapply(list(pI, .Machine$double.xmin),format, digits = 7), "\n")###################################################### code chunk number 29: sessionInfo###################################################toLatex(sessionInfo())###################################################### code chunk number 30: finalizing###################################################options(op.orig)。
平面相变热传导问题等效热容法的有限元解
大 连 理 工 大 学 学 报 J ournal of Dalian University of Technol ogy
Vol . 40, No. 1 J an. 2 0 0 0
文章编号 : 1000-8608( 2000) 01-0045-04
平面相变热传导问题等效热容法的有限元解
收稿日期 : 1998-12-01; 修订日期 : 1999-07-14 基金项目 : 国家杰出青年科学基金资助项目 ( 19525206) 作者简介 : 李海梅 ( 1969~ ) , 女 , 博士生 ; 顾元宪 ( 1954~ ) , 男 , 教授 , 博士生导师 ; 申长雨 ( 1963~ ) , 男 , 教授 , 博士生导师 .
表1 不同网格下求得的相界面位置及其与解析解的误差
T ab. 1 The numer ical results of phase inter face under different finite mesh and their er ror comparison with the analytical r esults
( 9) +
$t
C - (1- N ) K Tt $t 式中: N 是时间差分系数 , 0 < N< 1.
3] 引入焓 H 、 熵 S 两个物理量[ 2、 : T cpl ( T - T f) + $ H , H = T cp dT = f cpl ( T - T f) ,
( 10)
∫
T ≥ Tf T < Tf ( 11) 图1 相界面的求解示意图
解, 40单元的最大误差在 6% 以内. 时间差分格式的
变化 对 程序 计算 结 果影 响 很小 , 对 于 本 题, N = 0. 500, 0. 250, 0. 667的计算结果一样. 图 4中的温度 曲线与文献[ 6] 中的解析解曲线吻合得非常好 .
- 1、下载文档前请自行甄别文档内容的完整性,平台不提供额外的编辑、内容补充、找答案等附加服务。
- 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
- 3、如文档侵犯您的权益,请联系客服反馈,我们会尽快为您处理(人工客服工作时间:9:00-18:30)。
a r X i v :c o n d -m a t /0005060v 1 [c o n d -m a t .s t a t -m e c h ] 2 M a y 2000Finite size effects on the phase diagram of a binary mixture confined betweencompeting wallsM.M¨u ller and K.BinderInstitut f¨u r Physik,Johannes Gutenberg Universit¨a t Mainz,D-55099Mainz,Staudinger Weg 7,Germanyand E.V.Albano INIFTA,Universidad de la Plata,CC16Suc.4,1900La Plata,Argentina A symmetrical binary mixture AB that exhibits a critical temperature T cb of phase separation into an A-rich and a B-rich phase in the bulk is considered in a geometry confined between two parallel plates a distance D apart.It is assumed that one wall preferentially attracts A while the other wall preferentially attracts B with the same strength (”competing walls”).In the limit D →∞,one then may have a wetting transition of first order at a temperature T w ,from which prewetting lines extend into the one phase region both of the A-rich and the B-rich phase.It is discussed how this phase diagram gets distorted due to the finiteness of D :the phase transition at T cb immediately disappears for D <∞due to finite size rounding,and the phase diagram instead exhibit two two-phase coexistence regions in a temperature range T trip <T <T c 1=T c 2.In the limit D →∞T c 1,T c 2become the prewetting critical points and T trip →T w .For small enough D it may occur that at a tricritical value D t the temperatures T c 1=T c 2and T trip merge,and then for D <D t there is a single unmixing critical point as in the bulk but with T c (D )near T w .As an example,for the experimentally relevant case of a polymer mixture a phase diagram with two unmixing critical points is calculated explicitly from self-consistent field methods.[will be published in Physica A 279(No.1-4)(2000)pp.188–194.]I.INTRODUCTION Although the finite size effects on phase transitions in thin films have been studied since a long time [1–10],only during the last decade it was discovered [11–18]that in ferromagnetic Ising films with surface fields of different sign but of the same strength ±H 1(”competing walls”)novel types of phase transitions occur:namely,a phase transition occurs for zero bulk field H from a state with an interface freely fluctuating in the center of the thin film to a state where the interface is bound either to the lower or to the upper wall confining the film,which then acquires a nonzero (positive or negative)magnetization.This interface localization-delocalization transition at T c (D )may be either second [12–17]or first order [13,17,18],and for film thicknesses D →∞the transition temperature T c (D )does not converge towards the bulk critical temperature T cb as usual [1–5,8–10],but rather towards the wetting transition temperature T w (H 1)[12–18].Now it is well-known that,in general,wetting transitions [19]can be second or first order [20].Thus it is plausible that also the interface localization-delocalization transition can be either second or first order.However,recently it was shown [17]that also in cases where the wetting transition is first order,the transition T c (D )may be second order for small enough thickness (D <D t )and become first order only for D >D t .Thus the transition at T c (D t )is the finite-thickness analog of a wetting tricritical point [13–17].All this work [11–18]has only considered the case H =0,however.It is well known of course,that for D →∞,in the case of a first-order wetting transition at T w (H 1<0)there exists at T =T pre (H,H 1)a prewetting transition [19,20],where the distance of the interface bound to the wall discontinuously jumps from a smaller value to a larger value.While the analog of this prewetting transition in thin films has been studied occasionally for the case of ”capillary condensation”(where on both walls fields act that have the same sign)[7,10],it is only in the present work that the effect of prewetting phenomena on the interface localization-delocalization transition is considered [21].The physical systems that we have in mind are not magnets,of course,but rather binary (A,B)mixtures:as is well known,in an Ising model context the ”magnetization”simply translates into the relative concentration φif one component (A,say),and the field H translates into the chemical potential difference ∆µbetween the species (for simplicity we deal here with perfectly symmetric mixtures for which the bulk critical concentration is φcb =0.5).However,one important aspect of binary mixtures is that physically it is a density of an extensive thermodynamic variable (namely φ)that is the fixed independent thermodynamic variable,rather than the intensive variable ∆µ.As we shall see below,this fact has important consequences for the phase diagram of confined mixtures:the typical situation is that one encounters two successive lateral phase separation transitions!In Section2we elaborate these ideas by a qualitative discussion of the phase diagrams{both in the space of intensive variables(∆µ,T)and in the space(φ,T)}and of the corresponding physical state of the confined mixture.Section3 exemplifies these considerations by presenting a specific calculation for a symmetric polymer mixture,treated within a self-consistentfield framework[21].There are numerous experimental studies of confined polymer mixtures[22]and these systems might be convenient for testing our predictions.Finally section4summarizes our conclusions.II.QUALITATIVE PHASE DIAGRAMS OF CONFINED BINARY MIXTURESWe assume here a binary mixture confined by”competing”walls in the sense that one wall attracts species A with the same strength as the opposite wall attracts species B,and consider the case offirst-order wetting.Then the topology of the phase diagram can be estimated from the qualitative considerations as shown in the left part of Fig.1: In the space(T,∆µ),bulk(D→∞)phase separation occurs for∆µ=0,and the walls are incompletely wet for T<T w but wet for T w<T<T cb.From the point T=T w,∆µ=0there extend two(first-order)prewetting lines, which end at the prewetting critical points T cp.These prewetting transitions correspond to singularities of the surface free energies associated with the lower and upper wall confining the mixture(Fig.2).Due to the special symmetries chosen for our model,both the wetting transitions for the lower and upper wall coincide,and the prewetting critical temperatures are also the same.There is a mirror symmetry of the phase diagrams around the line∆µ=0(upper part)orφ=0.5(lower part),respectively.Forfinite thickness D it may happen,as demonstrated by Monte Carlo simulations for Ising models with enhanced exchange interactions near the walls[17],that a tricritical point at D=D t,T c(D t),∆µ=0(andφ=0.5)occurs. For D<D t then there exists a single critical point at T=T c(D),∆µ=0(andφ=1/2),there is no remnant of the prewetting phenomena left,and the phase diagram both in the(T,∆µ)plane as well as in the(T,φ)plane looks qualitatively exactly like in the bulk three-dimensional case.Of course,we do expect aflatter shape near the critical point due to the occurence of the two-dimensional Ising exponents{φcoex−1/2∝(1−T/T c(D))1/8rather thanφcoex−1/2∝(1−T/T cb)βwithβ≈0.325in the bulk[1–5,9]}.But the situation differs very much for D>D t (middle and right part of Fig.1).In the semi-grand-canonical ensemble(∆µfixed),one experiences a singlefirst order transition for−∆µc(D)<∆µ<∆µc(D)where±∆µc(D)is the chemical potential reached for T=T c(D) along the remnants of the prewetting lines.In the canonical ensemble(φfixed),we encounter a singlefirst-order phase transition only ifφ=φtrip=1wetting transition as in the Ising model[17],rather onefinds alwaysfirst order wetting behavior except close to the critical point of the bulk[10].As in our previous work[10,18]we consider a situation where the wetting transition temperature T w lies in the strong segregation regime,for which the self-consistentfield theory is accurate.The technical aspects of this approach are explained in detail elsewhere[10,21,24].Fig.3shows that for a typical choice of parameters indeed a phase diagram of the type of Fig.1,right part,is reproduced.Note that the self-consistentfield theory near the critical points T c1(D),T c2(D)implies mean-field like behavior,φcoex−φc1∝(1−T/T c1(D))1/2,rather than yielding the expected two-dimensional Ising exponent[10].But for large molecular weight this Ising like critical region is expected to be rather narrow[22],and thus we consider Fig.3as a useful hint for the phase diagram to be searched for in the experiments.IV.CONCLUSIONSIn this paper we have considered the problem of phase-separating binary mixtures confined between”competing walls”and have shown by qualitative considerations(Fig.1.)and self–consistentfield calculations(Fig.3.)that the phase diagram has either critical points andfirst order regions coexisting at a triple line or a single critical point resulting from the merging of these two critical points at the tricritical thickness D t.In previous work treating the case D>D t,only the case∆µ=0in the semi-grandcanonical ensemble was studied[17,18],which in the(T,φ)plane means that one cools the system atφ=φtrip=0.5and then a singlefirst order transition(Fig.2,left part)occurs: thus the existence of the two critical points was not previously discussed.Of course,in reality one will have to abandon the special symmetry assumptions used in Figs.1-3,allowing for asymmetric mixtures,differences in strength of the wall forces,etc,and thus the space of parameters to be considered gets much enlarged.However,as long as one works in the subspace where the wetting transition temperatures T w of both walls are the same,the phase diagrams still should have the topology of Fig.1,only the mirror symmetry around∆µ=∆µcoex(T)orφ=φcb is lost,and thus in generalφtrip will differ fromφcb.Also T c1and T c2will differ. Of course,in the most general case one must allow also for T w1=T w2,different wetting transition temperatures of both walls.One can also considerfirst order wetting at one wall and second order wetting at the other.A description of the phase diagrams for these more complicated cases is a challenging task for future work. Acknowledgements:[1]M.E.Fisher,in Critical Phenomena,M.S.Green,ed.(Academic Press,New York1971)[2]K.Binder and P.C.Hohenberg,Phys.Rev.B6(1972)3461;ibid B9(1974)2194.[3]K.Binder,Thin Solid Films20(1974)374.[4]M.E.Fisher and H.Au-Yang,Physica101(1980)255;H.Au-Yang and M.E.Fisher,Phys.Rev.B21(1980)3956.[5]M.E.Fisher and H.Nakanishi,J.Chem.Phys.75(1981)5857;H.Nakanishi and M.E.Fisher,J.Chem.Phys.78(1983)3279.[6]D.Sornette,Phys.Rev.B31(1985)4672;J.P.Desideri and D.Sornette,J.Phys.(France)49(1988)1411.[7]D.Nicolaides and R.Evans,Phys.Rev.B39(1989)9336.[8]K.Binder and ndau,J.Chem.Phys.96(1992)1444.[9]Y.Rouault,J.Baschnagel and K.Binder,J.Stat.Phys.80(1995)1009.[10]M.M¨u ller and K.Binder,Macromolecules31(1998)8323.[11]E.V.Albano,K.Binder,D.W.Heermann and W.Paul,Surf.Sci.223(1989)151.[12]A.O.Parry and R.Evans,Phys.Rev.Lett.64(1990)439.[13]M.R.Swift,A.L.Owczarek and J.O.Indekeu,Europhys.Lett.14(1991)475.[14]A.O.Parry and R.Evans,Physica A181(1992)250.[15]K.Binder,ndau,and A.M.Ferrenberg,Phys.Rev.Lett.74(1995)298;Phys.Rev.E51(1995)2823.[16]K.Binder,R.Evans,ndau,and A.M.Ferrenberg,Phys.Rev.E53(1996)5023.[17]A.M.Ferrenberg,ndau,and K.Binder,Phys.Rev.E58(1998)3353.[18]K.Binder and M.M¨u ller,Macromol.Symposia(1999,in press)[19]J.W.Cahn,J.Chem.Phys.66(1977)3667.[20]S.Dietrich,in:Phase Transitions and Critical Phenomena,vol.12,C.Domb and J.L.Lebowitz,eds.(Academic Press,New York1988)p.1[21]For a much more detailed account of our calculations,see M.M¨u ller,E.V.Albano,and K.Binder,preprint ”A symmetricpolymer blend confined into a thin film:interplay between wetting behavior and phase diagram”.[22]K.Binder,Adv.Polymer Sci.138(1999)1;A.Budkowski,Adv.Polymer Sci.(1999)[23]T.Kerle,J.Klein,and K.Binder,Phys.Rev.Lett.77(1996)1318;Europ.Phys.J.B 7(1999)401.[24]M.M¨u ller,Macromolecular Theory and Simulations 8(1999)343.o oT cTw T cpT cbD=D<D t T c TD>>D t T c T D>D tT cb T w T cpTT FIG.1.Qualitative phase diagrams of a bulk system (thickness D =∞)confined by two walls at whichcompeting ”fields”act,as well as corresponding phase diagrams of thin films of various thicknesses D =D <D t(left part),D >>D t (middle part)and D >D t (right part).Upper part of the panel presents the phase diagramsin the space of two intensive variables,lower part chooses as abscissa instead the concentration φ,a density of anextensive variable.Characteristic temperatures shown are the bulk critical (T cb ),wetting (T w )and prewettingcritical (T cp )temperatures,while in the thin film critical temperatures T c (D )occur at ∆µ=0and φ=φcb =1/2only for D <D t ,while for D >D t one has for ∆µ=0and φtrip =1/2rather a triple point of three phasecoexistence.However,two critical points T c 1=T c 2=T c (D )still occur,but at concentrations φ1c ,φ2c that movetowards the concentrations of the prewetting critical points as D →∞.For further explanations cf.text.000000000000000000000000000000111111111111111111111111111111111111111111111000000000000000000000000000000000000000000000000000000000000000000000000000000000000111111111111111111111111111111111111111111111111111111111111111111111111111111111111000000000000000000000000000000000000000000000000000000000000000000000000000000000000111111111111111111111111111111111111111111111111111111111111111111111111111111111111000000000000000000000000000000000000000000000111111111111111111111111111111111111111111111B-rich A-rich111111111111111000000000000000111111111111111wall wall B-rich 000000000000000000000000000000111111111111111111111111111111000000000000000000000000000000111111111111111111111111111111000000000000000000000000000000111111111111111111111111111111000000000000000000000000000000111111111111111111111111111111000000000000000000000000000000000000000000000000000000000000000000000000000000001111111111111111111111111111111111111111111111111111111111111111111111111111111100000000000000000000000000001111111111111111111111111111000000000000000000000000000000000000000000000000000000000000000000000000000111111111111111111111111111111111111111111111111111111111111111111111111111000000000000000000000000000011111111111111111111111111110000000000000000000000000000000000000000000001111111111111111111111111111111111111111111110000000000000000000000000000000000000000000001111111111111111111111111111111111111111111110000000000000000000000000000000000000000000000000000000011111111111111111111111111111111111111111111111111111111000000000000000000000000000011111111111111111111111111110000000000000000000000000000111111111111111111111111111100000000000000000000000000000000000000000000000000000000111111111111111111111111111111111111111111111111111111110000000000000000000000000000000000000000000000000000000000000000011111111111111111111111111111111111111111111111111111111111111111000000000000000000000000000000000000000000000000000000001111111111111111111111111111111111111111111111111111111100000000000000000000000000000000000000000000000000000000000000000000001111111111111111111111111111111111111111111111111111111111111111111111000000000000000000000000000000000000000000000111111111111111111111111111111111111111111111interfaceφ<φ<φ1trip T(D)<T<Tcb cb T T trip +T<T tripT<T φ=φtrip trip A-rich phase B-rich phase phase A-rich B-rich phase FIG.2.Qualitative explanation of the different phases that occur in a binary mixture confined between com-peting walls.Already at T >T cb there are A-rich and B-rich enrichment layers at the respective walls,having a finite thickness ξb (the bulk correlation length of concentration fluctuations),so for D >>2ξb the system is still disordered in the center,while in the region of temperatures where D and ξb are comparable a rounded transition to a already structure with one A-B interface at T <T cb occurs.For φ=φtrip the position of this interface is just in the middle of the thin film,while for φ1<φ<φtrip the location of the interface reflects the asymmetry of composition.For φ=φtrip one encounters a single transition at T =T trip ,where the localization of the interface at the walls requires lateral phase separation between A-rich and B-rich phases of equal amounts.For φ1<φ<φtrip one encounters two lateral phase separations:the first one is a coexistence between the phase with delocalized interface (in the center of the film as T →T +trip )with a phase where the interface is localized at the upper wall (the B-rich phase).In a second transition at T <T trip the phase with delocalized interface disappears in favor of the phase with the interface bound to the lower wall (the A-rich phase).Note that the amount of this phase must be less,to comply with the lever rule.−0.2−0.100.10.2∆µ/k B T 0.080.10.12T ~1/χN00.20.40.60.81φ0.080.10.12T ~1/χN FIG.3.Phase diagram of a symmetric polymer mixture in a thin film with antisymmetric boundary fields calculated in the framework of the self-consistent field theory for Gaussian chain molecules.Three film thicknesses D =0.5R e ,0.9R e and 2.6R e are shown (R e :end-to-end distance of the molecules).The smallest film thickness corresponds to the situation D <D t while films of thickness D =0.9R e and 2.6R e exhibit two critical points.Panel (a)displays the phase diagrams in the (T ∼1χNand composition φ.Note that the critical temperature of the bulk is at 1/χN =0.5.。