电力系统负荷预测及方法外文翻译.doc

合集下载

电力系统常用英文词汇

电力系统常用英文词汇

电力系统常用英文词汇电力专业英语词汇(较全)1、元件设备三绕组变压器three-column transformer ThrClnTrans 双绕组变压器double-column transformer DblClmnTrans电容器Capacitor并联电容器shunt capacitor电抗器Reactor母线Busbar输电线TransmissionLine发电厂power plant断路器Breaker刀闸(隔离开关)Isolator分接头tap电动机motor2状态参数有功active power无功reactive power电流current容量capacity电压voltage档位tap position有功损耗reactive loss无功损耗active loss空载损耗no-load loss铁损iron loss铜损copper loss空载电流no-load current阻抗impedance正序阻抗positive sequence impedance负序阻抗negative sequence impedance零序阻抗zero sequence impedance无功负载reactive load 或者QLoad有功负载: active load PLoad遥测YC(telemetering) 遥信YX励磁电流(转子电流)magnetizing current定子stator功角power-angle 上限:upper limit下限lower limit并列的apposable高压: high voltage低压low voltage中压middle voltage电力系统 power system发电机 generator励磁 excitation励磁器 excitor电压 voltage电流 current母线 bus变压器 transformer升压变压器 step-up transformer高压侧 high side输电系统 power transmission system输电线 transmission line固定串联电容补偿fixed series capacitor compensation 稳定stability电压稳定 voltage stability功角稳定 angle stability暂态稳定 transient stability电厂 power plant能量输送 power transfer交流 AC装机容量 installed capacity电网 power system落点 drop point开关站 switch station双回同杆并架 double-circuit lines on the same tower 变电站transformer substation补偿度 degree of compensation高抗 high voltage shunt reactor无功补偿 reactive power compensation故障 fault调节 regulation裕度 magin三相故障 three phase fault故障切除时间 fault clearing time极限切除时间 critical clearing time切机 generator triping 高顶值 high limited value强行励磁 reinforced excitation线路补偿器 LDC(line drop compensation)机端 generator terminal静态 static (state)动态 dynamic (state)单机无穷大系统 one machine - infinity bus system机端电压控制 AVR 功角 power angle有功功率 active power无功功率 reactive power功率因数 power factor无功电流 reactive current下降特性 droop characteristics斜率 slope额定 rating变比 ratio参考值 reference value电压互感器 PT分接头 tap下降率 droop rate仿真分析 simulation analysis传递函数 transfer function框图 block diagram受端 receive-side裕度 margin同步 synchronization失去同步 loss of synchronization阻尼 damping摇摆 swing保护断路器 circuit breaker电阻resistance电抗reactance阻抗impedance电导conductance电纳susceptance导纳admittance电感inductance电容: capacitanceAGC Automatic Generation Control自动发电控制AMR Automatic Message Recording 自动抄表ASS Automatic Synchronized System 自动准同期装置ATS Automatic Transform System 厂用电源快速切换装置AVR Automatic Voltage Regulator 自动电压调节器BCS Burner Control System 燃烧器控制系统BMS Burner Management System 燃烧器管理系统CCS Coordinated Control System 协调控制系统CRMS Control Room Management System 控制室管理系统CRT Cathode Ray Tube 阴极射线管DAS Data Acquisition System 数据采集与处理系统DCS Distributed Control System 分散控制系统DDC Direct Digital Control 直接数字控制系统DEH Digital Electronic Hydraulic Control 数字电液(调节系统) DPU Distributed Processing Unit 分布式处理单元EMS Energy Management System 能量管理系统ETS Emergency Trip System 汽轮机紧急跳闸系统EWS Engineering Working Station 工程师工作站FA Feeder Automation 馈线自动化FCS Field bus Control System 现场总线控制系统FSS Fuel Safety System 燃料安全系统FSSS Furnace Safeguard Supervisory System 炉膛安全监控系统GIS Gas Insulated Switchgear 气体绝缘开关设备GPS Global Position System 全球定位系统HCS Hierarchical Control System 分级控制系统LCD Liquid Crystal Display 液晶显示屏LCP Local Control Panel 就地控制柜MCC Motor Control Center 电动机马达控制中心MCS Modulating Control System 模拟量控制系统MEH Micro Electro Hydraulic Control System 给水泵汽轮机电液控制系统MIS Management Information System 管理信息系统NCS Net Control System 网络监控系统OIS Operator Interface Station 操作员接口站OMS Outage Management System 停电管理系统PID Proportion Integration Differentiation 比例积分微分PIO Process Output 过程输入输出通道PLC Programmable Logical Controller 可编程逻辑控制器PSS Power System Stabilizator 电力系统稳定器SCADA Supervisory Control And Data Acquisition 数据采集与监控系统SCC Supervisory Computer Control 监督控制系统SCS Sequence Control System 顺序(程序)控制系统SIS Supervisory Information System 监控信息系统 TDCS TDC Total Direct Digital Control 集散控制系统TSI Turbine Supervisory Instrumentation 汽轮机监测仪表UPS Uninterrupted Power Supply 不间断供电标准的机组数据显示(Standard Measurement And Display Data)负载电流百分比显示 Percentage of Current load(%)单相/三相电压 Voltage by One/Three Phase (Volt.)每相电流 Current by Phase (AMP)千伏安Apparent Power (KVA) 中线电流Neutral Current (N Amp)功率因数 Power Factor (PF)频率 Frequency(HZ)千瓦 Active Power (KW)千阀 Reactive Power (KVAr)最高/低电压及电流 Max/Min. Current and Voltage输出千瓦/兆瓦小时 Output kWh/MWh运行转速 Running RPM机组运行正常 Normal Running超速故障停机 Overspeed Shutdowns低油压故障停机 Low Oil Pressure Shutdowns高水温故障停机 High Coolant Temperature Shutdowns起动失败停机 Fail to Start Shutdowns冷却水温度表 Coolant Temperature Gauge机油油压表 Oil Pressure Gauge电瓶电压表 Battery Voltage Meter机组运行小时表 Genset Running Hour Meter怠速-快速运行选择键 Idle Run – Normal Run Selector Switch 运行-停机-摇控启动选择键Local Run-Stop-Remote Starting Selector Switch 其它故障显示及输入 Other Common Fault Alarm Display and电力行波词汇行波travelling wave模糊神经网络fuzzy-neural network神经网络neural network模糊控制fuzzy control研究方向 research direction电力系统the electrical power system大容量发电机组large capacity generating set输电距离electricity transmission超高压输电线super voltage transmission power line投运commissioning行波保护Traveling wave protection自适应控制方法adaptive control process动作速度speed of action行波信号travelling wave signal输电线路故障transmission line malfunction子系统subsystem暂态行波transient state travelling wave偏移量side-play mount电压electric voltage附加系统add-ons system波形waveform工频power frequency延迟变换delayed transformation延迟时间delay time减法运算subtraction相减运算additive operation求和器summator模糊规则fuzzy rule参数值parameter values可靠动作action message等值波阻抗equivalent value wave impedance 附加网络additional network反传算法backpropagation algorithm隶属函数membership function模糊规则fuzzy rule模糊推理fuzzy reasoning模糊推理矩阵fuzzy reasoning matrix样本集合 sample set感应定律law of electromagnetic induction 励磁 excitation励磁器 magnetizing exciter励磁器 exciter恒定励磁器constant exciter励磁器激振器exciter励磁电流magnetizing current强行励磁 reinforced excitation励磁调节器 excitation regulator无功伏安 volt-ampere reactive无功伏安时volt-ampere-hour reactive三绕组变压器:three-column transformer ThrClnTrans 双绕组变压器:double-column transformer DblClmnTrans 电容器:Capacitor并联电容器:shunt capacitor电抗器:Reactor母线:Busbar输电线:TransmissionLine发电厂:power plant断路器:Breaker刀闸(隔离开关):Isolator分接头:tap电动机:motor(2)状态参数有功:active power无功:reactive power电流:current容量:capacity电压:voltage档位:tap position有功损耗:reactive loss无功损耗:active loss功率因数:power-factor功率:power功角:power-angle电压等级:voltage grade空载损耗:no-load loss铁损:iron loss铜损:copper loss空载电流:no-load current阻抗:impedance正序阻抗:positive sequence impedance 负序阻抗:negativesequence impedance 零序阻抗:zero sequence impedance 电阻:resistor电抗:reactance电导:conductance电纳:susceptance无功负载:reactive load 或者QLoad有功负载: active load 或者PLoad遥测:YC(telemetering)遥信:YX励磁电流(转子电流):magnetizing current 定子:stator功角:power-angle上限:upper limit下限:lower limit并列的:apposable高压: high voltage低压:low voltage中压:middle voltage电力系统 power system发电机 generator励磁 excitation励磁器 excitor电压 voltage电流 current母线 bus变压器 transformer升压变压器 step-up transformer高压侧 high side输电系统 power transmission system输电线 transmission line固定串联电容补偿fixed series capacitor compensation 稳定电压稳定 voltage stability功角稳定 angle stability暂态稳定 transient stability电厂 power plant能量输送 power transfer交流 AC装机容量 installed capacity电网 power system落点 drop point开关站 switch station双回同杆并架 double-circuit lines on the same tower 变电站transformer substation补偿度 degree of compensation高抗 high voltage shunt reactor无功补偿 reactive power compensation故障 fault调节 regulation裕度 magin三相故障 three phase fault故障切除时间 fault clearing time极限切除时间 critical clearing time切机 generator triping高顶值 high limited value强行励磁 reinforced excitation线路补偿器 LDC(line drop compensation)机端 generator terminal静态 static (state)动态 dynamic (state)单机无穷大系统 one machine - infinity bus system 机端电压控电抗 reactance电阻 resistance功角 power angle有功(功率) active power无功(功率) reactive power功率因数 power factor无功电流 reactive current下降特性 droop characteristics 斜率 slope额定 rating变比 ratio参考值 reference value电压互感器 PT分接头 tap下降率 droop rate仿真分析 simulation analysis传递函数 transfer function框图 block diagram受端 receive-side裕度 margin同步 synchronization失去同步 loss of synchronization 阻尼 damping摇摆 swing保护断路器 circuit breaker电阻:resistance电抗:reactance阻抗:impedance电导:conductance电纳:susceptance导纳:admittance电感:inductance电容: capacitanceAbsorber Circuit ——吸收电路AC/AC Frequency Converter——交交变频电路AC power control——交流电力控制AC Power Controller——交流调功电路AC Power Electronic Switch——交流电力电子开关Ac Voltage Controller——交流调压电路Asynchronous Modulation 异步调制Baker Clamping Circuit 贝克箝位电路Bi-directional Triode Thyristor 双向晶闸管Bipolar Junction Transistor 双极结型晶体管(BJT)Boost-Buck Chopper 升降压斩波电路Boost Chopper 升压斩波电路Boost Converter 升压变换器Bridge Reversible Chopper 桥式可逆斩波电路Buck Chopper 降压斩波电路Buck Converter 降压变换器Commutation 换流Conduction Angle 导通角Constant Voltage Constant Frequency 恒压恒频(CVCF)Continuous Conduction--CCM (电流)连续模式Control Circuit控制电路Cuk Circuit——CUK斩波电路Current Reversible Chopper 电流可逆斩波电路Current Source Type Inverter--CSTI 电流(源)型逆变电路Cycloconvertor 周波变流器DC-AC-DC Converter 直交直电路DC Chopping 直流斩波DC Chopping Circuit 直流斩波电路DC-DC Converter 直流-直流变换器Device Commutation 器件换流Direct Current Control 直接电流控制Discontinuous Conduction mode (电流)断续模式displacement factor 位移因数distortion power 畸变功率double end converter 双端电路driving circuit 驱动电路electrical isolation 电气隔离fast acting fuse 快速熔断器fast recovery diode 快恢复二极管fast recovery epitaxial diodes 快恢复外延二极管fast switching thyristor 快速晶闸管field controlled thyristor 场控晶闸管flyback converter 反激电流forced commutation 强迫换流forward converter 正激电路frequency converter 变频器full bridge converter 全桥电路full bridge rectifier 全桥整流电路full wave rectifier 全波整流电路fundamental factor 基波因数gate turn-off thyristor——GTO 可关断晶闸管general purpose diode 普通二极管giant transistor——GTR 电力晶体管half bridge converter 半桥电路hard switching 硬开关high voltage IC 高压集成电路hysteresis comparison 带环比较方式indirect current control 间接电流控制indirect DC-DC converter 直接电流变换电路insulated-gate bipolar transistor---IGBT绝缘栅双极晶体管intelligent power module---IPM 智能功率模块integrated gate-commutated thyristor---IGCT集成门极换流晶闸管inversion 逆变latching effect 擎住效应leakage inductance 漏感light triggered thyristo---LTT光控晶闸管line commutation 电网换流load commutation 负载换流loop current 环流。

电力系统负荷预测及方法(外文翻译)

电力系统负荷预测及方法(外文翻译)

电力系统负荷预测及方法(外文翻译)Power system load forecasting methods and characteristics of Abstract: The load forecasting in power system planning and operation play an important role, with obvious economic benefits, in essence, the electricity load forecasting market demand forecast. In this paper, a systematic description and analysis of a variety of load forecasting methods and characteristics and that good load forecasting for power system has become an important means of modern management.Keywords: power system load forecasting electricity market construction Planning1.IntroductionLoad forecasting demand for electricity from a known starting to consider the political, economic, climate and other related factors, the future demand for electricity to make predictions. Load forecast includes two aspects: on the future demand (power) projections and future electricity consumption (energy) forecast.Electricity demand projections decision generation, transmission and distribution system, the sic of new Capacity; power generating equipment determine the type of prediction (.such as peaking units, base load units, etc}.Load forecasting purposes is to provide load conditions and the level of development, while identifying the various supply areas, each year planning for the power consumption for maximum power load and the load of planning the overall level of development of each plan year to determine the load composition.2. load forecasting methods and characteristics of2.1 Unit Consumption ActOutput of products in accordance with national arrangements, planning and electricity intensity value to determine electricity demand. Sub-Unit Consumption Act; Product Unit Consumption; and the value of Unit Consumption Act; two. The projection of load before the key is to determine the appropriate value of the product unit consumption or unitconsumption. Judging from China's actual situation, the general rule is the product unit consumption increased year by year, the output value unit consumption is declining. Unit consumption method advantages arc: The method is simple, short-torn load forecasting effective. Disadvantages arc: need to do a lot of painstaking research work, more general, it is difficult to reflect modern economic, political and climate conditions.2.2 Trend extrapolationWhen the power load in accordance with time-varying present same kind of upward or downward trend, and no obvious seasonal fluctuations, but also to find a suitable function curve to reflect this change in trend, you can use the time t as independent variables, timing value of y for the dependent variable to establish the trend model y = f (t). When the reason to believe that this trend will extend to the future, we assigned the value of the variable t need to, you can get the corresponding tune series of the future value of the moment. This is the trendextrapolation.Application of the trend extrapolation method has two assumptions: (1) assuming there is no step Change in load; (2)assume that the development of load factors also determine the future development of load and its condition is unchanged or changed little. Select the appropriate trend model is the application of the trend extrapolation an important part of pattern recognition method and finite difference method is to select the trend model arc two basic ways.A linear trend extrapolation forecasting method, the logarithmic trend forecasting method, quadratic curve trend forecasting method, exponential curve trend forecasting method, growth curve of the trend prediction method. Trend extrapolation method's advantages arc: only need to historical data, the amount of data required for less. The disadvantage is that: If a change in load will cause large errors.2.3 Elastic Coefficient MethodElasticity coefficient is the average growthrate of electricity consumption to GDP ratio of between, according to the gross domestic product growth rate of coefficient of elasticity to be planning with the end of the total electricity consumption. Modules of elasticity law is determined on power development from a macro with the relative speed of national economic development, which is a measure of national economic development and an important parameter in electricity demand. The advantages of this method arc: The method is simple, easy to calculate. Disadvantages arc: need to do a lot of detailed research work.2.4 Regression Analysis MethodRegression estimate is based on past history of load data, build up a mathematical analysis of the mathematical model. Of mathematical statistics regression analysis of the variables in statistical analysis of observational data in order to achieve load to predict the future. Regression model with a linear regression, multiple linear regression, nonlinear regression and other regressionprediction models. Among them, linear regression for the medium-torn toad forecast. Advantages arc: a higher prediction accuracy for the medium and the use of short-term forecasts. The disadvantage is that: (1) planning level it is difficult years of industrial and agricultural output statistics; (2) regression analysis can only be measured out the level of development of an integrated electricity load can not be measured out the power supply area of the loading level of development, thus can notbe the specific grid construction plan.2.5 Time Series AnalysisThe load is on the basis of historical data, trying to build a mathematical model, using this mathematical model to describe the power load on the one hand this random variable of statistical regularity of the change process; the other hand, the mathematical model based on the re-establishment of the mathematical expression of load forecasting type, to predict the future load. Time series are mainlyautoregressive AR (p), moving average MA (q) and self-regression and n3oving average ARMA (p, q) and so on. The advantages of these methods arc: the historical data required for less, work less. The disadvantage is that: There is no change in load factor to consider, only dedicated to the data fitting, the lack of regularity of treatment is only applicable to relatively uniform changes in the short-term load forecasting situation.2.6 Gray model methodGray prediction is a kind of a system containing uncertain factors to predict approach. Gray system theory based on the gray forecasting techniques may be limited circumstances in the data to identify the role of law within a certain period, the establishment of load forecasting models. Is divided into ordinary gray system model and optimization model for two kinds of gray.Ordinary gray prediction model is an exponential growth model, when the electric load in strict accordance with exponentiallygrowing, this method has high accuracy and required less sample data to calculate simple and testable etc.; drawback is that for a change in volatility The power load, the prediction error largo, does not meet actual needs. And the gray model optimization can have ups and downs of the original data sequence transformed into increased exponentially increasing regularity changes in sequence, greatly improving prediction accuracy and the gray model method of application. Gray Model Law applies to short-torn load forecast. Gray predicted advantages: smaller load data requirements, without regard to the distribution of laws and do not take into account trends, computing convenient, short-term forecasts of high precision, easy to test. Drawbacks: First, when the data the greater the degree of dispersion, namely, the greater the gray level data, prediction accuracy is worse; 2 is not very suitable for the long-term power system to push a number of years after the forecast.2.7 Delphi MethodThe Delphi method is based on the special knowledge of direct experience, research problems of judgment, a method for prediction of, also called experts investigation. Delphi method has feedback, anonymity and statistical characteristics. Delphi method advantage is:(1) can accelerate prediction speed and save prediction Cost; (2)can get different but valuable ideas and opinions; (3)suitable for long-term forecasts in historical data, insufficient or unpredictable factors is particularly applicable more. Detect is: (1)the load forecasting far points area may not reliable;(2)the expert opinions sometimes may not complete or impractical.2.8 Expert System ApproachExpert system prediction is stored in the database over the past tow years, even decades, the Hourly load and weather data analysis, which brings together experienced staff knowledge load forecasting, extract the relevant rules, according to certain rules, load prediction.Practice has proved that accurate load forecasting requires not only high-tech support, but also need to reconcile the experience and wisdom of mankind itself: Therefore, you need expert systems such technologies. Expert systems approach is a non-quantifiable human experience translated into a better way But experts systems analysis itself is a time-consuming process, and some complex factors (such as weather factors), even though aware of its load impact, ht}t to accurately and quantitatively determine their influence on the load area is also very difficult. Expert system for forecasting method suitable for medium and long-term load forecast. The advantages of this method: (1)can bring together multiple expert knowledge and experience to maximize the ability of experts; (2) possession of data, information and mort factors to consider a more comprehensive and beneficial to arrive at mart accurate conclusions. The disadvantage is that: (1)do not have the self-learning ability, subjectto the knowledge stored in the database limits the total; (2) pairs of unexpected incidents and poor adaptability to changing conditions2.9 Neural Network MethodNeural network (ANN, Artificial Neural Network) forecasting techniques to mimic the human brain to do intelligent processing, a large number of non-structural. non-deterministic laws of adaptive function. ANN used in short-term load forecasting and long-term load forecast than that applied to be mart appropriate. Because short-term load changes can be regarded as a stationary random process. And long-term load forecasting may be due to political, economic and other major fuming point leading to a mathematical model-based damage. Advantages arc:(1) to mimic the human brain, intelligence processing; (2}a large number of non-structural. non-adaptive function of the accuracy of the law; (3)with the information memory, self-learning, knowledge, reasoningand optimization of computing features. The disadvantage is that:(1) the determination of the initial value can not take advantage of existing system information, easily trapped in local minimum of the state; (2) neural network learning process is usually slow, poor adaptability to sudden events.2.10 Optimum Combination Forecasting MethodOptimal combination has two meanings: First, several forecasting methods from the results obtained by selecting the appropriate a0cight in the weighted average; 2 refers to the comparison of several prediction methods, choose the best or the degree of preparation and the standard deviation of the smallest prediction model forecast. For the combined forecasting method must also noted that the combined forecast is a single forecasting model can not completely correct to describe the changes of the amount predicted to play a role. One can fully reflect the actual law of development of the model predictions agree well with the combination forecasting method than predictedgood results. This method has the advantage: To optimize the combination of a wide range of information on a single prediction model, consider the impact of information is also mart comprehensive, so it can effectively improve the prediction. The disadvantage is that: (1) the weight is difficult to determine; (2) all possible factors that play a role in the future, all included in the model, to a certain extent, limit the prediction accuracy improved.2.11 Wavelet analysis and forecasting techniquesWavelet analysis is a time-domain-frequency domain analysis method, it is in the time domain and frequency domain at the same time has good localization properties, and can automatically adjust according to the signal sampling frequency of high and low density, it is cast' to capture and analysis of weak signals and signal, images of any small parts. The advantage is: Can the different frequency components gradually refined using a sampling step, which can be gathered in any of the details of the signal, especially for singular signal is very sensitive tothe treatment well or mutation weak signals, their goal is to a signal information into wavelet coefficients, which can easily be dealt with, storage, transmission, analysis or for the reconstruction of the original signal. These advantages determine the wavelet analyses can be effectively applied to load forecasting issues.3. ConclusionLoad forecasting is the electric power system scheduling, real-time control, operation plan and development planning, the premise is a grid dispatching departments and planning departments must have the basic information. Improve load forecasting technology level, be helpful for program management, reasonable arrangement of the electricity grid operation mode for the maintenance plan and the crew, to section coal, fuel-efficient and reduce generating cost, be helpful for formulate rational power construction planning of the power system, improve the economic benefit andsocial benefit. Therefore, the load forecast has become a power system management modernization realization of the important content.电力系统负荷预测及方法摘要:负荷预测在电力系统规划和运行方面发挥的重要作用,具有明显的经济效益,负荷预测实质上是对电力市场需求的预测。

电力系统中英文翻译

电力系统中英文翻译

LINE PROTECTION WITH DISTANCE RELAYSDistance relaying should be considered when overcurrent relaying is too slow or is not selective. Distance relays are generally used for phase-fault primary and back-up protection on subtransmission lines, and on transmission lines where high-speed automatic reclosing is not necessary to maintain stability and where the short time delay for end-zone faults can be tolerated. Overcurrent relays have been used generally for ground-fault primary and back-up protection, but there is a growing trend toward distance relays for ground faults also.Single-step distance relays are used for phase-fault back-up protection at the terminals of generators. Also, single-step distance relays might be used with advantage for back-up protection at power-transformer tanks, but at the present such protection is generally provided by inverse-time overcurrent relays.Distance relays are preferred to overcurrent relays because they are not nearly so much affected by changes in short-circuit-current magnitude as overcurrent relays are, and , hence , are much less affected by changes in generating capacity and in system configuration. This is because, distance relays achieve selectivity on the basis of impedance rather than current.THE CHOICE BETWEEN IMPEDANCE, REACTANCE, OR MHOBecause ground resistance can be so variable, a ground distance relay must be practically unaffected by large variations in fault resistance. Consequently, reactance relays are generally preferred for ground relaying.For phase-fault relaying, each type has certain advantages and disadvantages. For very short line sections, the reactance type is preferred for the reason that more of theline can be protected at high speed. This is because the reactance relay is practically unaffected by arc resistance which may be large compared with the line impedance, as described elsewhere in this chapter. On the other hand, reactance-type distance relays at certain locations in a system are the most likely to operate undesirably on severe synchronizing-power surges unless additional relay equipment is provided to prevent such operation.The mho type is best suited for phase-fault relaying for longer lines, and particularly where severe synchronizing-power surges may occur. It is the least likely to require additional equipment to prevent tripping on synchronizing-power surges. When mho relaying is adjusted to protect any given line section, its operating characteristic encloses the least space on the R-X diagram, which means that it will be least affected by abnormal system conditions other than line faults; in other words, it is the most selective of all distance relays. Because the mho relay is affected by arc resistance more than any other type, it is applied to longer lines. The fact that it combines both the directional and the distance-measuring functions in one unit with one contact makes it very reliable.The impedance relay is better suited for phase-fault relaying for lines of moderate length than for either very short or very long lines. Arcs affect an impedance relay more than a reactance relay but less than a mho relay. Synchronizing-power surges affect an impedance relay less than a reactance relay but more than a mho relay. If an impedance-relay characteristic is offset, so as to make it a modified• relay, it can be made to resemble either a reactance relay or a mho relay but it will always require a separate directional unit.There is no sharp dividing line between areas of application where one or another type of distance relay is best suited. Actually, there is much overlapping of these areas. Also, changes that are made in systems, such as the addition of terminals to a line, can change the type of relay best suited to a particular location. Consequently, to realizethe fullest capabilities of distance relaying, one should use the type best suited for each application. In some cases much better selectivity can be obtained between relays of the same type, but, if relays are used that are best suited to each line, different types on adjacent lines have no appreciable adverse effect on selectivity. THE ADJUSTMENT OF DISTANCE RELAYSPhase distance relays are adjusted on the basis of the positive-phase-sequence impedance between the relay location and the fault location beyond which operation of a given relay unit should stop. Ground distance relays are adjusted in the same way, although some types may respond to the zero-phase-sequence impedance. This impedance, or the corresponding distance, is called the "reach" of the relay or unit. For purposes of rough approximation, it is customary to assume an average positive-phase-sequence-reactance value of about 0.8 ohm per mile for open transmission-line construction, and to neglect resistance. Accurate data are available in textbooks devoted to power-system analysis.To convert primary impedance to a secondary value for use in adjusting a phase or ground distance relay, the following formula is used:where the CT ratio is the ratio of the high-voltage phase current to the relay phase current, and the VT ratio is the ratio of the high-voltage phase-to-phase voltage to the relay phase-to-phase voltage–all under balanced three-phase conditions. Thus, for a 50-mile, 138-kv line with 600/5 wye-connected CT’s, the secondary positive-phase-sequence reactance is aboutIt is the practice to adjust the first, or high-speed, zone of distance relays to reach to80% to 90% of the length of a two-ended line or to 80% to 90% of the distance to the nearest terminal of a multiterminal line. There is no time-delay adjustment for this unit.The principal purpose of the second-zone unit of a distance relay is to provide protection for the rest of the line beyond the reach of the first-zone unit. It should be adjusted so that it will be able to operate even for arcing faults at the end of the line. To do this, the unit must reach beyond the end of the line. Even if arcing faults did not have to be considered, one would have to take into account an underreaching tendency because of the effect of intermediate current sources, and of errors in: (1) the data on which adjustments are based, (2) the current and voltage transformers, and (3) the relays. It is customary to try to have the second-zone unit reach to at least 20% of an adjoining line section; the farther this can be extended into the adjoining line section, the more leeway is allowed in the reach of the third-zone unit of the next line-section back that must be selective with this second-zone unit.The maximum value of the second-zone reach also has a limit. Under conditions of maximum overreach, the second-zone reach should be short enough to be selective with the second-zone units of distance relays on the shortest adjoining line sections, as illustrated in Fig. 1. Transient overreach need not be considered with relays having a high ratio of reset to pickup because the transient that causes overreach will have expired before the second-zone tripping time. However, if the ratio of reset to pickup is low, the second-zone unit must be set either (1) with a reach short enough so that its overreach will not extend beyond the reach of the first-zone unit of the adjoining linesection under the same conditions, or (2) with a time delay long enough to be selective with the second-zone time of the adjoining section, as shown in Fig. 2. In this connection, any underreaching tendencies of the relays on the adjoining line sections must be taken into account. When an adjoining line is so short that it is impossible to get the required selectivity on the basis of react, it becomes necessary to increase the time delay, as illustrated in Fig. 2. Otherwise, the time delay of the second-zone unit should be long enough to provide selectivity with the slowest of (1) bus-differential relays of the bus at the other end of the line(2)transformer-differential relays of transformers on the bus at the other end of the line,or (3) line relays of adjoining line sections. The interrupting time of the circuit breakers of these various elements will also affect the second-zone time. This second-zone time is normally about 0.2 second to 0.5 second.The third-zone unit provides back-up protection for faults in adjoining line sections.So far as possible, its reach should extend beyond the end of the longest adjoining line section under the conditions that cause the maximum amount of underreach, namely, arcs and intermediate current sources. Figure 3 shows a normal back-up characteristic. The third-zone time delay is usually about 0.4 second to 1.0 second. To reach beyond the end of a long adjoining line and still be selective with the relays of a short line, it may be necessary to get this selectivity with additional time delay, as in Fig. 4.THE EFFECT OF ARCS ON DISTANCE-RELAY OPERATIONThe critical arc location is just short of the point on a line at which a distance relay's operation changes from high-speed to intermediate time or from intermediate time to back-up time. We are concerned with the possibility that an arc within the high-speed zone will make the relay operate in intermediate time, that an arc within the intermediate zone will make the relay operate in back-up time, or that an arc within the back-up zone will prevent relay operation completely. In other words, the effect of an arc may be to cause a distance relay to underreach.For an arc just short of the end of the first- or high-speed zone, it is the initial characteristic of the arc that concerns us. A distance relay's first-zone unit is so fast that, if the impedance is such that the unit can operate immediately when the arc is struck, it will do so before the arc can stretch appreciably and thereby increase itsresistance. Therefore, we can calculate the arc characteristic for a length equal to the distance between conductors for phase-to-phase faults, or across an insulator string for phase-to-ground faults. On the other hand, for arcs in the intermediate-time or back-up zones, the effect of wind stretching the arc should be considered, and then the operating time for which the relay is adjusted has an important bearing on the outcome.Tending to offset the longer time an arc has to stretch in the wind when it is in the intermediate or back-up zones is the fact that, the farther an arcing fault is from a relay, the less will its effect be on the relay's operation. In other words, the more line impedance there is between the relay and the fault, the less change there will be in the total impedance when the arc resistance is added. On the other hand, the farther away an arc is, the higher its apparent resistance will be because the current contribution from the relay end of the line will be smaller, as considered later.A small reduction in the high-speed-zone reach because of an arc is objectionable, but it can be tolerated if necessary. One can always use a reactance-type or modified-impendance type distance relay to minimize such reduction. The intermediate-zone reach must not be reduced by an arc to the point at which relays of the next line back will not be selective; of course, they too will be affected by the arc, but not so much. Reactance-type or modified-impendance-type distance relays are useful here also for assuring the minimum reduction in second-zone reach. Figure 5 shows how an impedance or mho characteristic can be offset to minimize its susceptibility to an arc. One can also help the situation by making the second-zone reach as long as possible so that a certain amount of reach reduction by an arc is permissible. Conventional relays do not use the reactance unit for the back-up zone; instead, they use either an impedance unit, a modified-impendance unit, or a mho unit. If failure of the back-up unit to operate because of an arc extended by the wind is a problem, the modified-impendance unit can be used or the mho–or "starting"–unitcharacteristic can also be shifted to make its operation less affected by arc resistance. The low-reset characteristic of some types of distance relay is advantageous in preventing reset as the wind stretches out an arc.Although an arc itself is practically all resistance, it may have a capacitive-reactance or an inductive-reactance component when viewed from the end of a line where the relays are. The impedance of an arc (ZA) has the appearance:where I1 = the complex expression for the current flowing into the arc from the end of the line where the relays under consideration are.I2= the complex expression for the current flowing into the arc from the other end of the line.R A = the arc resistance with current (I1 + I2) flowing into it.Of more practical significance is the fact that, as shown by the equation, the arc resistance will appear to be higher than it actually is, and it may be very much higher. After the other end of the line trips, the arc resistance will be higher because the arccurrent will be lower. However, its appearance to the relays will no longer be magnified, because I2 will be zero. Whether its resistance will appear to the relays to be higher or lower than before will depend on the relative and actual magnitudes of the currents before and after the distant breaker opens.输电线路的距离保护在过电流保护灵敏度低或选择性差时,应当考虑采用距离保护。

电力负荷和故障外文文献翻译、中英文翻译、外文翻译

电力负荷和故障外文文献翻译、中英文翻译、外文翻译

附录Electrical Loads and FaultsPart1 Electrical LoadsThe electrical load devices used in industry, in our homes and in commercial buildings are very important parts of electrical power systems. The load of any system performs a function which involves power conversion. A load converts one form of energy to another. An electrical load converts electrical energy to some other form of energy, such as heat, light or mechanical energy. Electrical loads may be classified according to the function which they perform or by the electrical characteristics which they exhibit.In order to plan for power system load requirements, it is necessary to understand the electrical characteristics of all the loads connected to the power system. The types of power supplies and distribution systems which a building uses are determined by the load characteristics. All loads may be considered as either resistive, inductive, capacitive, or a combination of these. We should be aware of the effects which various types of loads will have on the power system. The nature of alternating current causes certain electrical circuit properties to exist.One primary factor which affects the electrical power system is the presence of inductive loads. These are mainly electric motors. To counteract the inductive effects, utility companies use power-factor corrective capacitors as part of the power system design. capacitor units are located at substations to improve the power factor of the system. The inductive effect, therefore, increases the cost of a power system and reduces the actual amount of power which is converted to another from of energy.Part2 faults and its DamageEach year new designs of power equipment bring about increased reliability of operation . Nevertheless, equipment failures and interference by outside sources occasionally result in faults on electric power systems . On the occurrence of a faults, current and voltage conditions become abnormal, the delivery of power from the generating stations to the loads may be unsatisfactory over a considerable area , and if the faulty equipment is not promptly disconnected from the remainder of the system ,damage may result to other pieces ofoperating equipment .A fault is the unintentional or intentional connecting together of two or more conductors which ordinarily operate with a difference of potential between them . The connection between the conductors may be by physical metallic contact or it may be through an arc .At the fault ,the voltage between the two parts is reduced to zero in the case of metal-to-metal contacts, or to a very low value in the case the connection is through an arc . Currents of abnormally high magnitude flow through the network to the point of fault . These short-circuit currents will usually be much greater than the designed thermal ability of the conductors in the lines or machines feeding the fault . The resultant rise in temperature may cause damage by the annealing of conductors and by the charring of insulation . In the period during which the fault is permitted to exist ,the voltage on the system in the near vicinity of the fault will be so low that utilization equipment will be inoperative .It is apparent that the power system designer must anticipate points at which faults may occur ,be able to calculate conductors that exist during a faults, and provide equipment properly adjusted to open the switches necessary to disconnect the faults equipment from the remainder of the system .Ordinarily it is desirable that no other switches on the system are opened , as such behavior would result in unnecessary modification of the system .Part 3 overloadA distinction must be made between a fault and an overload .An overload only that loads greater than the designed values have been imposed on system . Under such a circumstance the voltage at the overload point may be low ,but not zero . This undervoltage condition may extend for some distance beyond the overload point into the remainder of the system .The current in the overload equipment are high and may exceed the thermal design limits . Nevertheless , such current are substantially lower than in the case of a fault . Service frequently may be maintained , but at below-standard voltage.Overload are rather common occurrences in homes. For example, a housewife might plug five waffle irons into the kitchen circuit during a neighborhood party . Such an overload , if permitted to continue , would cause heating of the power center and might eventually start a fire . To prevent such trouble , residential circuits are protected by fuses or circuit breakers which open quickly when currents above specified values persist . Distribution transformersare sometimes overload as customers instant more and more appliances. The continuous monitoring of distribution circuits is necessary to be certain that transformer sizes are increased as load grows.Part4 Various FaultsFaults of many types and cause may appear on power system . Many of our homes have seen frayed lamp cords which permitted the two conductions of the cord to come in contact with each other . When this is occurs , there is a resulting flash , and if breakers or fuse equipment functions properly ,the circuit is opened.Overload lines , for the most part, are constructed of bare conductors . These are sometimes accidentally brought together by action of wind ,sleet, trees, cranes, airplanes, or damage to supporting structures .Overvoltages due to lightning or switching may cause flashover of supporting or from conductors to conductors .Contamination on insulators sometimes results in flashover even during normal voltage conditions.The conductors of underground cables are separated from each other and from ground by solid insulation, which may be oil-impregnated paper or a plastic such as polyethylene . These materials undergo some deterioration with age ,particularly if overload on the cables have resulted in their operation at elevated temperature . Any small void present in the body of the insulating material will result in ionization of the gas contained therein , the products of which react unfavorably with the insulation .Deterioration of the insulation may result in failure of the material to retain its insulating properties , and short circuits will develop between the cable conductors . The possibility of cable failure is increased if lighting or switching produces transient voltage of abnormally high values between the conductors .Transformer failures may be the result of insulation deterioration combined with overvoltages due to lightning or switching transients transients. Short circuits due to insulation failure between adjacent turns of the same winding may result from suddenly applied overvoltages. Major insulation may fail, permitting arcs to ge established between primary and secondary windings or between a winding and grounded metal parts such as the core or tank.Generators may fail due to breakdown of the insulation between adjacent turns in the same slot, resulting in a short circuit in a single turn of the generator. Insulation breakdown may also occur between one of the windings and the grounded steel structure in which the coils areembedded. Breakdown between different windings lying in the same slot results in short-circuiting extensive sections of machine.Balanced three-phase faults, like balanced three-phase loads, may be handled on a line to-neutral basis or on an equivalent single-phase basis. Problems may be solved either in terms of volts, amperes, and ohms. The handling of faults on single-phase lines is of course identical to the method of handling three-phase faults on an equivalent single-phase basis.Part5 permanent faults and temporary faultsFaults may be classified as permanent or temporary. Permanent faults are those in which insulation failure or structure failure produces damage that makes operation of the equipment impossible and requires repairs to be made. Temporary faults are those which overhead lines frequently are of this nature. High winds may cause two or more conductors may continue as long as the line remains energized. However, if automatic equipment can be brought into operation to deenergize the line quickly, little physical damage may result and the line may be restored to service as soon as the are is extinguished. Arcs across insulators due to overvoltages from lightning or switching transients usually can be cleared by automatic circuit-breaker operation before significant structure damage occurs.Because of this characteristic of faults on lines, many companies operate following a procedure known as high-speed reclosing. On the occurrence of a fault, the line is promptly deenergized by opening the circuit breakers at each end of the line. The breakers remain open long enough for the arc to clear, and then reclose automatically. In many instances service is restored in a fraction of a second. Of course, if structure damage has occurred and the fault persists, it is necessary for the breakers to reopen and lock open.电力负荷和故障一.电力负荷用于工业、家庭、商业建筑的负载设备是电力系统中十分重要的组成部分。

电力负荷预测中英文对照外文翻译文献

电力负荷预测中英文对照外文翻译文献

中英文资料翻译基于改进的灰色预测模型的电力负荷预测[摘要]尽管灰色预测模型已经被成功地运用在很多领域,但是文献显示其性能仍能被提高。

为此,本文为短期负荷预测提出了一个GM(1,1)—关于改进的遗传算法(GM(1,1)-IGA)。

由于传统的GM(1,1)预测模型是不准确的而且参数α的值是恒定的,为了解决这个问题并提高短期负荷预测的准确性,改进的十进制编码遗传算法(GA)适用于探求灰色模型GM(1,1)的最佳α值。

并且,本文还提出了单点线性算术交叉法,它能极大地改善交叉和变异的速度。

最后,用一个日负荷预测的例子来比较GM(1,1)-IGA模型和传统的GM(1,1)模型,结果显示GM(1,1)-IGA拥有更好地准确性和实用性。

关键词:短期的负荷预测,灰色系统,遗传算法,单点线性算术交叉法第一章绪论日峰值负荷预测对电力系统的经济,可靠和安全战略都起着非常重要的作用。

特别是用于每日用电量的短期负荷预测(STLF)决定着发动机运行,维修,功率互换和发电和配电任务的调度。

短期负荷预测(STLF)旨在预测数分钟,数小时,数天或者数周时期内的电力负荷。

从一个小时到数天以上不等时间范围的短期负荷预测的准确性对每一个电力单位的运行效率有着重要的影响,因为许多运行决策,比如:合理的发电量计划,发动机运行,燃料采购计划表,还有系统安全评估,都是依据这些预测[]1。

传统的负荷预测模型被2,3,4。

通常,这些模型对于日常的短期负荷预测是有效的,分为时间序列模型和回归模型[]5,6,7。

此外,由于它们的复杂性,为了获但是对于那些特别的日子就会产生不准确的结果[]得比较满意的结果需要大量的计算工作。

8,9,10,主要是模型的不确定性和信息不完整的灰色系统理论最早是由邓聚龙提出来的[]分析,对系统研究条件的分析,预测以及决策。

灰色系统让每一个随机变量作为一个在某一特定范围内变化的灰色量。

它不依赖于统计学方法来处理灰色量。

它直接处理原始数据,来寻找数据内在的规律[]11。

电力负荷预测中英文外文翻译文献

电力负荷预测中英文外文翻译文献

中英文资料外文翻译文献基于改进的灰色预测模型的电力负荷预测[摘要]尽管灰色预测模型已经被成功地运用在很多领域,但是文献显示其性能仍能被提高。

为此,本文为短期负荷预测提出了一个GM(1,1)—关于改进的遗传算法(GM(1,1)-IGA)。

由于传统的GM(1,1)预测模型是不准确的而且参数α的值是恒定的,为了解决这个问题并提高短期负荷预测的准确性,改进的十进制编码遗传算法(GA)适用于探求灰色模型GM(1,1)的最佳α值。

并且,本文还提出了单点线性算术交叉法,它能极大地改善交叉和变异的速度。

最后,用一个日负荷预测的例子来比较GM(1,1)-IGA模型和传统的GM(1,1)模型,结果显示GM(1,1)-IGA拥有更好地准确性和实用性。

关键词:短期的负荷预测,灰色系统,遗传算法,单点线性算术交叉法第一章绪论日峰值负荷预测对电力系统的经济,可靠和安全战略都起着非常重要的作用。

特别是用于每日用电量的短期负荷预测(STLF)决定着发动机运行,维修,功率互换和发电和配电任务的调度。

短期负荷预测(STLF)旨在预测数分钟,数小时,数天或者数周时期内的电力负荷。

从一个小时到数天以上不等时间范围的短期负荷预测的准确性对每一个电力单位的运行效率有着重要的影响,因为许多运行决策,比如:合理的发电量计划,发动机运行,燃料采购计划表,还有系统安全评估,都是依据这些预测[]1。

传统的负荷预测模型被2,3,4。

通常,这些模型对于日常的短期负荷预测是有效的,分为时间序列模型和回归模型[]5,6,7。

此外,由于它们的复杂性,为了获但是对于那些特别的日子就会产生不准确的结果[]得比较满意的结果需要大量的计算工作。

8,9,10,主要是模型的不确定性和信息不完整的灰色系统理论最早是由邓聚龙提出来的[]分析,对系统研究条件的分析,预测以及决策。

灰色系统让每一个随机变量作为一个在某一特定范围内变化的灰色量。

它不依赖于统计学方法来处理灰色量。

它直接处理原始数据,来寻找数据内在的规律[]11。

电力系统自动化毕业论文中英文资料外文翻译

电力系统自动化毕业论文中英文资料外文翻译

毕业设计(论文)外文资料翻译专业名称:电力系统自动化英文资料:INDUCTION MOTOR STARTING METHODSAbstract -Many methods can be used to start large AC induction motors. Choices such as full voltage, reduced voltage either by autotransformer or Wyes - Delta, a soft starter, or usage of an adjustable speed drive can all have potential advantages and trade offs. Reduced voltage starting can lower the starting torque and help prevent damage to the load. Additionally, power factor correction capacitors can be used to reduce the current, but care must be taken to size them properly. Usage of the wrong capacitors can lead to significant damage. Choosing the proper starting method for a motor will include an analysis of the power system as well as the starting load to ensure that the motor is designed to deliver the needed performance while minimizing its cost. This paper will examine the most common starting methods and their recommended applications.I. INTRODUCTIONThere are several general methods of starting induction motors: full voltage, reduced voltage, wyes-delta, and part winding types. The reduced voltage type can include solid state starters, adjustable frequency drives, and autotransformers. These, along with the full voltage, or across the line starting, give the purchaser a large variety of automotives when it comes to specifying the motor to be used in a given application. Each method has its own benefits, as well as performance trade offs. Proper selection will involve a thorough investigation of any power system constraints, the load to be accelerated and the overall cost of the equipment.In order for the load to be accelerated, the motor must generate greater torque than the load requirement. In general there are three points of interest on the motor's speed-torque curve. The first is locked-rotor torque (LRT) which is the minimum torque which the motor will develop at rest for all angular positions of the rotor. The second is pull-up torque (PUT) which is defined as the minimum torque developed by the motor during the period of acceleration from rest to the speed at which breakdown torque occurs. The last is the breakdown torque (BDT) which is defined as the maximum torque which the motor will develop. If any of these points are below the required load curve, then the motor will not start.The time it takes for the motor to accelerate the load is dependent on the inertia of the load and the margin between the torque of the motor and the load curve, sometimes called accelerating torque. In general, the longer the time it takes for the motor to accelerate the load, the more heat that will be generated in the rotor bars, shorting ring and the stator winding. This heat leads to additional stresses in these parts and can have an impaction motor life.II. FULL VOLTAGEThe full voltage starting method, also known as across the line starting, is the easiest method to employ, has the lowest equipment costs, and is the most reliable. This method utilizes a control to close a contactor and apply full line voltage to the motor terminals. This method will allow the motor to generate its highest starting torque and provide the shortest acceleration times.This method also puts the highest strain on the power system due to the high starting currents that can be typically six to seven times the normal full load current of the motor. If the motor is on a weak power system, the sudden high power draw can cause a temporary voltage drop, not only at the motor terminals, but the entire power bus feeding the starting motor. This voltage drop will cause a drop in the starting torque of the motor, and a drop in the torque of any other motor running on the power bus. The torque developed by an induction motor varies roughly as the square of the applied voltage. Therefore, depending on the amount of voltage drop, motors running on this weak power bus could stall. In addition, many control systems monitor under voltage conditions, a second potential problem that could take a running motor offline during a full voltage start. Besides electrical variation of the power bus, a potential physical disadvantage of an across the line starting is the sudden loading seen by the driven equipment. This shock loading due to transient torques which can exceed 600% of the locked rotor torque can increase the wear on the equipment, or even cause a catastrophic failure if the load can not handle the torques generated by the motor during staring.A. Capacitors and StartingInduction motors typically have very low power factor during starting and as a result have very large reactive power draw. See Fig. 2. This effect on the system can be reduced by adding capacitors to the motor during starting.The large reactive currents required by the motor lag the applied voltage by 90 electrical degrees. This reactive power doesn't create any measurable output, but is rather the energy required for the motor to function. The product of the applied system voltage and this reactive power component can be measured in V ARS (volt-ampere reactive). The capacitors act to supply a current that leads the applied voltage by 90 electrical degrees. The leading currents supplied by the capacitors cancel the laggingcurrent demanded by the motor, reducing the amount of reactive power required to be drawn from the power system.To avoid over voltage and motor damage, great care should be used to make sure that the capacitors are removed as the motor reaches rated speed, or in the event of a loss of power so that the motor will not go into a generator mode with the magnetizing currents provided from the capacitors. This will be expanded on in the next section and in the appendix.B. Power Factor CorrectionCapacitors can also be left permanently connected to raise the full load power factor. When used in this manner they are called power factor correction capacitors. The capacitors should never be sized larger than the magnetizing current of the motor unless they can be disconnected from the motor in the event of a power loss.The addition of capacitors will change the effective open circuit time constant of the motor. The time constant indicates the time required for remaining voltage in the motor to decay to 36.8% of rated voltage after the loss of power. This is typically one to three seconds without capacitors.With capacitors connected to the leads of the motor, the capacitors can continue to supply magnetizing current after the power to the motor has been disconnected. This is indicated by a longer time constant for the system. If the motor is driving a high inertia load, the motor can change over to generator action with the magnetizingCurrent from the capacitors and the shaft driven by the load. This can result in the voltage at the motor terminals actually rising to nearly 50% of rated voltage in some cases. If the power is reconnected before this voltage decays severe transients can be created which can cause significant switching currents and torques that can severely damage the motor and the driven equipment. An example of this phenomenon is outlined in the appendix.Ⅲ. REDUCED VOLTAGEEach of the reduced voltage methods are intended to reduce the impact of motor starting current on the power system by controlling the voltage that the motor sees atthe terminals. It is very important to know the characteristics of the load to be started when considering any form of reduced voltage starting. The motor manufacturer will need to have the speed torque curve and the inertia of the driven equipment when they validate their design. The curve can be built from an initial, or break away torque, as few as four other data points through the speed range, and the full speed torque for the starting condition. A centrifugal or square curve can be assumed in many cases, but there are some applications where this would be problematic. An example would be screw compressors which have a much higher torque requirement at lower speeds than the more common centrifugal or fan load. See Fig. 3. By understanding the details of the load to be started the manufacturer can make sure that the motor will be able to generate sufficient torque to start the load, with the starting method that is chosen.A. AutotransformerThe motor leads are connected to the lower voltage side of the transformer. The most common taps that are used are 80%, 65%, and 50%. At 50% voltage the current on the primary is 25% of the full voltage locked rotor amps. The motor is started with this reduced voltage, and then after a pre-set condition is reached the connection is switched to line voltage. This condition could be a preset time, current level, bus volts, or motor speed. The change over can be done in either a closed circuit transition, or an open circuit transition method. In the open circuit method the connection to the voltage is severed as it is changed from the reduced voltage to the line level. Care should be used to make sure that there will not be problems from transients due to the switching. This potential problem can be eliminated by using the closed circuit transition. With the closed circuit method there is a continuousVoltage applied to the motor. Another benefit with the autotransformer starting is in possible lower vibration and noise levels during starting.Since the torque generated by the motor will vary as the square of the applied voltage, great care should be taken to make sure that there will be sufficient accelerating torque available from the motor. A speed torque curve for the driven equipment along with the inertia should be used to verify the design of the motor. A good rule of thumb is to have a minimum of 10% of the rated full load torque of the motor as a margin at all points of the curve.Additionally, the acceleration time should be evaluated to make sure that the motor has sufficient thermal capacity to handle the heat generated due to the longeracceleration time.B. Solid State or Soft StartingThese devices utilize silicon controlled rectifiers or Scars. By controlling the firing angle of the SCR the voltage that the device produces can be controlled during the starting of the motor by limiting the flow of power for only part of the duration of the sine wave.The most widely used type of soft starter is the current limiting type. A current limit of 175% to 500% of full load current is programmed in to the device. It then will ramp up the voltage applied to the motor until it reaches the limit value, and will then hold that current as the motor accelerates.Tachometers can be used with solid state starters to control acceleration time. Voltage output is adjusted as required by the starter controller to provide a constant rate of acceleration.The same precautions in regards to starting torque should be followed for the soft starters as with the other reduced voltage starting methods. Another problem due to the firing angle of the SCR is that the motor could experience harmonic oscillating torques. Depending on the driven equipment, this could lead to exciting the natural frequency of the system.C. Adjustable Frequency DrivesThis type of device gives the greatest overall control and flexibility in starting induction motors giving the most torque for an amount of current. It is also the most costly.The drive varies not only the voltage level, but also the frequency, to allow the motor to operate on a constant volt per hertz level. This allows the motor to generate full load torque throughout a large speed range, up to 10:1. During starting, 150% of rated current is typical.This allows a significant reduction in the power required to start a load and reduces the heat generated in the motor, all of which add up to greater efficiency. Usage of the AFD also can allow a smaller motor to be applied due to the significant increase of torque available lower in the speed range. The motor should still be sizedlarger than the required horsepower of the load to be driven. The AFD allows a great degree of control in the acceleration of the load that is not as readily available with the other types of reduced voltage starting methods.The greatest drawback of the AFD is in the cost relative to the other methods. Drives are the most costly to employ and may also require specific motor designs to be used. Based on the output signal of the drive, filtered or unfiltered, the motor could require additional construction features. These construction features include insulated bearings, shaft grounding brushes, and insulated couplings due to potential shaft current from common mode voltage. Without these features, shaft currents, which circulate through the shaft to the bearing, through the motor frame and back, create arcing in the bearings that lead to premature bearing failure, this potential for arcing needs to be considered when applying a motor/drive package in a hazardous environment, Division2/Zone2.An additional construction feature of a motor used on an AFD may require is an upgraded insulation system on the motor windings. An unfiltered output signal from a drive can create harmonic voltage spikes in the motor, stressing the insulation of the motor windings.It is important to note that the features described pertain to motors which will be started and run on an AFD. If the drive is only used for starting the motor, these features may not be necessary. Consult with the motor manufacturer for application specific requirements.D. Primary Resistor or Reactor StartingThis method uses either a series resistor or reactor bank to be placed in the circuit with the motor. Resistor starting is more frequently used for smaller motors.When the motor is started, the resistor bank limits the flow of inrush current and provides for a voltage drop at the motor terminals. The resistors can be selected to provide voltage reductions up to 50%. As the motor comes up to speed, it develops a counter EMF (electro-magnetic field) that opposes the voltage applied to the motor. This further limits the inrush currents. As the inrush current diminishes, so does t>e voltage drop across the resistor bank allowing the torque generated by the motor to increase. At a predetermined time a device will short across the resistors and open the starting contactor effectively removing the resistor bank from the circuit. This provides for a closed transition and eliminates the concerns due to switchingtransients.Reactors will tend to oppose any sudden changes in current and therefore act to limit the current during starting. They will remain shorted after starting and provide a closed transition to line voltage.E .Star delta StartingThis approach started with the induction motor, the structure of each phase of the terminal are placed in the motor terminal box. This allows the motor star connection in the initial startup, and then re-connected into a triangle run. The initial start time when the voltage is reduced to the original star connection, the starting current and starting torque by 2 / 3. Depending on the application, the motor switch to the triangle in the rotational speed of between 50% and the maximum speed. Must be noted that the same problems, including the previously mentioned switch method, if the open circuit method, the transition may be a transient problem. This method is often used in less than 600V motor, the rated voltage 2.3kV and higher are not suitable for star delta motor start method.Ⅴ. INCREMENT TYPEThe first starting types that we have discussed have deal with the way the energy is applied to the motor. The next type deals with different ways the motor can be physically changed to deal with starting issues.Part WindingWith this method the stator of the motor is designed in such a way that it is made up of two separate windings. The most common method is known as the half winding method. As the name suggests, the stator is made up of two identical balanced windings. A special starter is configured so that full voltage can be applied to one half of the winding, and then after a short delay, to the second half. This method can reduce the starting current by 50 to 60%, but also the starting torque. One drawback to this method is that the motor heating on the first step of the operation is greater than that normally encountered on across-the-line start. Therefore the elapsed time on the first step of the part winding start should be minimized. This method also increases the magnetic noise of the motor during the first step.IV .ConclusionThere are many ways asynchronous motor starting, according to the constraints of power systems, equipment costs, load the boot device to select the best method. From the device point of view, was the first full-pressure launch the cheapest way, but it may increase the cost efficiency in the use of, or the power supply system in the region can not meet their needs. Effective way to alleviate the buck starts the power supply system, but at the expense of the cost of starting torque.These methods may also lead to increased motor sizes have led to produce the required load torque. Inverter can be eliminated by the above two shortcomings, but requires an additional increase in equipment costs. Understand the limitations of the application, and drives the starting torque and speed, allowing you for your application to determine the best overall configuration.英文资料翻译:异步电动机起动的方法摘要:大容量的交流异步电动机有多种启动方法。

电力专业英语英文文献翻译报告

电力专业英语英文文献翻译报告

电力专业英语英文文献翻译报告Page 1.The Production of Electrical Energy(电能生产)1 English textFrom reference 1Should the generation be not adequate to balance the load demand, it is imperative that one of following alternatives be considered for keeping the system in operating condition:1. Staring fast peaking units,2. Load shedding for unimportant loads,3. generation rescheduling.It is apparent from the above that since the voltage specifications are not stringent, load frequency control is by far the most important in power system control.In order to understand the frequency control, consider a small step-increase in load. The initial distribution of the load increment is determined by the system simpedance; and the sistantaneous relative generator rotor positions. The energy required to supply the load increment is drawn from the kinetic energy of rotating machines. As a result, the system frequency drops. The distribution of load during this period among the various machines is determined by the inertias of the rotors of the generators partaking in process. This problem is stability analysis of the system.After the speed or frequency fall due to reduction in stored energy in the rotors has taken place, the drop is sensed by the governors and they divide the load increment between the machines as determined by the droops of the respective governor characterstics. Subsequently, secondary control restores the system frequency to its normal value by readjustingthe governor characteristics. Keywords:load frequency control From reference 2Modern power systems are so large that it is impossible to design a single central control system that would handle the overall control job. It is extremely useful take into account the weak links in the system and then apply control through decomposition. The demarcation of load frequency control and Mavar voltage control characteristics is one such decomposition. Geographical and functional decomposition are successfully applied to power systems and this leads to the concept of area control.A modern power system can be divided into several areas for load frequency control. Each control area fulfils the following:1.The area is a geographically contious portion of a large interconnected area, which adjusts its own generation to accommodate load changes within its precincts.2.Under normal conditions of operation, it changes bulk power with neighboring areas.3.Under abnormal conditions of operation, it may deviate from predetermined schedules and provide assistance to any neighboring control area in the system.4.It is expected, in addition, to partake with the other areas in the system in a suitable manner in the system frequency regulation.The rotors of all generators in a control area swing together for load change. Thus, a coherent group of generators within a geographical region may constitute a control area which is connected to other similar areas by weak tie lines.Keywords:areas load frequency controlFrom reference 3For plant loading schedules in thermal systems, load prediction up to two hours in advance is necessary while for unit commitment schedules prediction up to 24 hours is sufficient. Also, at all sations and control centers, short-time prediction is needed for storage and display of advance information. Based on this information, predictive security assessment of the system is made. This also helps to contain the rates of change of generator outputs within their permissible limits.For the implementation of economic scheduling of generation using digital computers, detailed estimates of the future load demands are essential in order to allow sufficient time for the calculation and implementation of the generator schedules. Whatever method is envisaged for the calculation of such economic schedules consistent with the security and spare requirements of the system, the schedules should be calculated every 15 or 30 minutes and each economic schedule should be a predictive one ,for at least about 30 minutes ahead of event. It is then obvious that the predictions are to be revised frequently in the light of any fresh information so as to minimize the estimation errors.Peak load demand forecasts are useful in determining the investment required for additional generating and transmission capacities required. Forecasts for planning require data extending over several previous years. Meaningful forecasts can be obtained with lead time of 3 to 5 years.Keywords:load predictionFrom reference 4In this method, the load is separated into two main components. The first component is a base load which is of fixed value and the second a variable component which is a functionof the weather conditions.Estimates can be made 24 hours ahead, using the weather forecast. The temperature base for weighting the effect of the predicated temperature on the load is the normal, mean temperature of the month. The normal, mean temperature of the month has zero weight. Similarly the change in consumers demand due to cloudy weather may be assumed to vary in direct proportion to the degree of cloudiness. This in turn may be expressed by an illumination index with fair, clear sky corresponding to zero weight.The base load is determined from past records. Proper weighting of the elements of the weather will be attained only after several trials. The method of prediction stabilizes after this trial period. It may be noted that the base loads for week days and weekend will generally be different for any hour.Using these base loads, a load estimate based on the best available weather forecast can be made using proper weighting of meteorological factors like temperature, cloudiness, wind velocity, etc.Keywords:proper weighting of the elements of the weather2 中文翻译及分析出自文献1:万一发电量不足以平衡负荷需求,要使电力系统处于运行状态,必须考虑采取以下选择方法中的一种:1、启动快速峰荷机组2、对不重要的用户实行拉闸断电3、重新制定发电计划从上述情况来看,电压技术的要求并不严格,目前为止负荷频率控制是电力系统控制中最重要的手段。

(完整版)电力系统外文英语文献资料

(完整版)电力系统外文英语文献资料

(完整版)电力系统外文英语文献资料Electric Power SystemElectrical power system refers to remove power and electric parts of the part,It includes substation, power station and distribution. The role of the power grid is connected power plants and users and with the minimum transmission and distribution network disturbance through transport power, with the highest efficiency and possibility will voltage and frequency of the power transmission to the user fixed .Grid can be divided into several levels based on the operating voltage transmission system, substructure, transmission system and distribution system, the highest level of voltage transmission system is ZhuWangJia or considered the high power grids. From the two aspects of function and operation, power can be roughly divided into two parts, the transmission system and substation. The farthest from the maximum output power and the power of the highest voltage grade usually through line to load. Secondary transmission usually refers to the transmission and distribution system is that part of the middle. If a plant is located in or near the load, it might have no power. It will be direct access to secondary transmission and distribution system. Secondary transmission system voltage grade transmission and distribution system between voltage level. Some systems only single second transmission voltage, but usually more than one. Distribution system is part of the power system and its retail service to users, commercial users and residents of some small industrial users. It is to maintain and in the correct voltage power to users responsible. In most of the system, Distribution system accounts for 35% of the total investment system President to 45%, andtotal loss of system of the half .More than 220kv voltage are usually referred to as Ultra high pressure, over 800kv called high pressure, ultra high voltage and high pressure have important advantages, For example, each route high capacity, reduce the power needed for the number of transmission. In as high voltage to transmission in order to save a conductor material seem desirable, however, must be aware that high voltage transmission can lead to transformer, switch equipment and other instruments of spending increases, so, for the voltage transmission to have certain restriction, allows it to specific circumstances in economic use. Although at present, power transmission most is through the exchange of HVDC transmission, and the growing interest in, mercury arc rectifier and brake flow pipe into the ac power generation and distribution that change for the high voltage dc transmission possible.Compared with the high-voltage dc high-voltage ac transmission has the following some advantages: (1) the communication with high energy; (2) substation of simple maintenance and communication cost is low; (3) ac voltage can easily and effectively raise or lower, it makes the power transmission and high pressure With safety voltage distribution HVDC transmission and high-voltage ac transmission has the following advantages: (1) it only need two phase conductors and ac transmission to three-phase conductors; (2) in the dc transmission impedance, no RongKang, phase shift and impact overvoltage; (3) due to the same load impedance, no dc voltage, and transfer of the transmission line voltage drop less communication lines, and for this reason dc transmission line voltage regulator has better properties; (4) in dc system withoutskin effect. Therefore, the entire section of route conductors are using; (5) for the same work, dc voltage potential stress than insulation. Therefore dc Wire need less insulation; (6) dc transmission line loss, corona to little interference lines of communication; (7) HVDC transmission without loss of dielectric, especially in cable transmission; (8) in dc system without stability and synchronization of trouble.A transmission and the second transmission lines terminated in substation or distribution substations, the substation and distribution substations, the equipment including power and instrument transformer and lightning arrester, with circuit breaker, isolating switch, capacitor set, bus and a substation control equipment, with relays for the control room of the equipment. Some of the equipment may include more transformer substations and some less, depending on their role in the operation. Some of the substation is manual and other is automatic. Power distribution system through the distribution substations. Some of them by many large capacity transformer feeders, large area to other minor power transformer capacity, only a near load control, sometimes only a doubly-fed wire feeders (single single variable substation)Now for economic concerns, three-phase three-wire type communication network is widely used, however, the power distribution, four lines using three-phase ac networks.Coal-fired power means of main power generating drive generators, if coal energy is used to produce is pushing the impeller, then generate steam force is called the fire. Use coal produces steam to promote the rotating impeller machine plant called coal-fired power plants. In the combustion process, the energy stored in the coal to heat released,then the energy can be transformed into the form within vapor. Steam into the impeller machine work transformed into electrical energy.Coal-fired power plants could fuel coal, oil and natural gas is. In coal-fired power plant, coal and coal into small pieces first through the break fast, and then put out. The coal conveyer from coal unloader point to crush, then break from coal, coal room to pile and thence to power. In most installations, according to the needs of coal is, Smash the coal storage place, no coal is through the adjustable coal to supply coal, the broken pieces of coal is according to the load changes to control needs. Through the broken into the chamber, the coal dust was in the second wind need enough air to ensure coal burning.In function, impeller machine is used to high temperature and high pressure steam energy into kinetic energy through the rotation, spin and convert electricity generator. Steam through and through a series of impeller machine parts, each of which consists of a set of stable blade, called the pipe mouth parts, even in the rotor blades of mobile Li called. In the mouth parts (channel by tube nozzle, the steam is accelerating formation) to high speed, and the fight in Li kinetic energy is transformed into the shaft. In fact, most of the steam generator is used for air is, there is spread into depression, steam turbine of low-pressure steam from the coagulation turbine, steam into the condenses into water, and finally the condensate water is to implement and circulation.In order to continuous cycle, these must be uninterrupted supply: (1) fuel; (2) the air (oxygen) to the fuel gas burning in the configuration is a must; (3) and condenser, condensed from the condensed water supply, sea and river to lake. Common coolingtower; (4) since water vapour in some places in circulation, will damage process of plenty Clean the supply.The steam power plant auxiliary system is running. For a thermal power plant, the main auxiliary system including water system, burning gas and exhaust systems, condensation system and fuel system. The main auxiliary system running in the water pump, condensation and booster pump, coal-fired power plants in the mill equipment. Other power plant auxiliary equipment including air compressors, water and cooling water system, lighting and heating systems, coal processing system. Auxiliary equipment operation is driven by motor, use some big output by mechanical drive pump and some of the impeller blades, machine drive out from the main use of water vaporimpeller machine. In coal-fired power plant auxiliary equipment, water supply pump and induced draft fan is the biggest need horsepower.Most of the auxiliary power generating unit volume increased significantly in recent years, the reason is required to reduce environment pollution equipment. Air quality control equipment, such as electrostatic precipitator, dust collection of flue gas desulfurization, often used in dust in the new coal-fired power plants, and in many already built in power plant, the natural drive or mechanical drive, fountain, cooling tower in a lake or cooling canal has been applied in coal-fired power plants and plants, where the heat release need to assist cooling system.In coal-fired power stations, some device is used to increase the thermal energy, they are (1) economizer and air preheater, they can reduce the heat loss; (2) water heater, he can increase the temperature of water into boiling water heaters; (3) they can increase and filter the thermal impeller.Coal-fired power plants usually requires a lot of coal and coal reservoirs, however the fuel system in power plant fuel handling equipment is very simple, and almost no fuel oil plants.The gas turbine power plants use gas turbine, where work is burning gas fluid. Although the gas turbine must burn more expensive oil or gas, but their low cost and time is short, and can quickly start, they are very applicable load power plant. The gas turbine burn gas can achieve 538 degrees Celsius in the condensing turbine, however, the temperature is lower, if gas turbine and condenser machine, can produce high thermal efficiency. In gas turbine turbine a combined cycle power plant. The gas through a gas turbine, steam generator heat recovery in there were used to generate vapor heat consumption. Water vapor and then through a heated turbine. Usually a steam turbine, and one to four gas turbine power plant, it must be rated output power.。

电力系统短期负荷预测

电力系统短期负荷预测

电力系统短期负荷预测POWER SYSTEM SHORT-TERM LOAD FORECASTING专业:电气工程及其自动化姓名:指导教师姓名:申请学位级别:学士论文提交日期:二零一六年十二月学位授予单位:天津科技大学摘要电力系统负荷预测是电力生产部门的重要工作之一.准确的负荷预测,可以合理安排机组启停,减少备用容量,合理安排检修计划及降低发电成本等.准确的预测,特别是短期负荷预测对提高电力经营主体的运行效益有直接的作用,对电力系统控制、运行和计划都有重要意义.因此,针对不同场合需要寻求有效的负荷预测方法来提高预测精度。

本文采用神经网络方法对电力系统短期负荷进行预测。

本文主要介绍了电力负荷预测的主要方法和神经网络的原理、结构,分析了反向传播算法,建立三层人工神经网络模型进行负荷预测,并编写相关程序。

与此同时采用最小二乘法进行对比,通过对最小二乘法多项式拟合原理的学习,建立模型编写相关程序。

通过算例对两种模型绝对误差、相对误差、拟合精度进行分析,同时比较它们训练时间,得出标准BP神经网络具有更好的精度优势但训练速度较慢。

最后针对标准BP神经网络训练速度慢、容易陷入局部最小值等缺点,对标准BP神经网络程序运用附加动量法进行修改,分析改进后网络的优点。

关键词:短期负荷预测标准BP神经网络最小二乘法附加动量法ABSTRACTPower system load forecasting is one of the most important work of the electricity production sector。

The accurate load forecasting can arrange unit start-stop, reduce the spare capacity, reasonable arrangement of the maintenance plan and reduce power cost,etc。

电力系统负荷预测的研究(开题报告,文献综述,论文,外文翻译)

电力系统负荷预测的研究(开题报告,文献综述,论文,外文翻译)

电力系统负荷预测的研究(开题报告,文献综述,论文,外文翻译)【毕业设计】电气自动化论文电力负荷预测方法的研究文献综述开题报告中期检查表外文翻译Q毕业设计(论文)文献综述电子与电气工程系2009级电气工程及其自动化陈AA09AAAAAA电力负荷预测方法的研究系别:年级专业:姓名:学号:题目名称:电力负荷预测方法的研究文献综述【内容摘要】:负荷预测是电力系统规划、计划、用电、调度等部门的基础工作。

讨论了年度负荷预测、月度负荷预测和短期负荷预测的特点、成熟方法,分析了负荷预测问题的各种解决方案,并指出未来的主要研究方向。

根据国内电力系统负荷预测的实践和国外的经验,对我国开展电力系统负荷预测工作提出了一些建议。

【关键词】:电力系统;负荷预测;模型;参数辨识电力负荷预测方法与应用一、概述电力工业是国民经济的基础工业。

随着我国产业结构完善和人民整体生活水平的改善,对电能的需求逐年加大,同时对电力质量的要求也越来越高,且由于电能生产和消费的同时性,对电网建设和布局提出了更高的要求。

电力负荷预测是电网规划建设的依据和基础。

随着电力工业在国民经济中扮演着越来越重要的角色,电力负荷的正确预测显得尤为重要。

电力负荷预测是指通过对电力系统负荷历史数据的分析和研究,运用统计学、数学、计算机、工程技术及经验分析等定性定量的方法,探索事物之间的内在联系和发展变化规律,对未来的负荷发展做出预先估计和推测。

电力负荷预测结果的准确与否直接关系到电力投资的效益,供电的可靠性,用电需求的正常发展,以及社会的经济效益和社会效益。

但要做到预测准确或较准确是很困难的,因为影响电力负荷预测的因素相当多,且由于各地区产业结构和人民生活水平不同,各具体因素对电力负荷预测的敏感度是不一样的,因而电力负荷预测具模糊性。

回顾我国“十五”期间的预测情况与实际发展情况是很有意义的。

基于“九五”期间国民经济和电力工业的发展状况,在全国电力供需趋于平衡的前提下,我国制定的“十五”规划对电力工业发展提出了“可持续发展”的要求:电力工业发展方式要从数量速度型向质量效益型转变,从以供给导向为主转向以需求导向为主,优化电力资源配置。

电力系统负荷预测综述

电力系统负荷预测综述

Electric Power Technology316电力系统负荷预测综述徐 健(国网江苏省电力有限公司苏州供电分公司,江苏 苏州 215004)摘要:本文对负荷预测的种类进行了全面的分析,并且对负荷预测的意义进行了详细的探讨。

在文章的最后对当前负荷预测的方法进行了调查。

关键词:电力系统;负荷;预测;综述改革开放以来,我国经济不断发展,人民生活水平不断提高,与此同时人民生活生产中所用到的用电设备也越来越多,因此人们对电力需求、电力服务质量、电力服务效率的要求也越来越高,电力企业急需快速发展来满足人们越来越高的需求。

因此有必要对电力系统的负荷预测进行研究,以保证电力企业能够得到更加长远的发展。

1 电荷预测种类在电力负荷预测工作进行的过程当中,由于涉及到的内容比较繁多,因此相当复杂。

从预测的目的以及所需的时间长度出发可以将负荷预测分为四大种类:分别是超短期负荷预测、短期负荷预测、中期负荷预测以及长期负荷预测。

下面分别对这四种预测进行探讨。

1.1 中长期负荷预测如果工作人员在进行负荷预测的过程当中,预计整个预测所需要的年限要大于等于十年,而且预测过程当中所需要用到的时间单位是一年来计算,那么就把该种负荷预测定义为长期负荷预测;如果整个负荷预测的年限大致为五年左右,预测过程当中所需要用到的时间单位为年的话,定义该预测方式为中期负荷预测;这两种类型的负荷预测工作进行的意义在于可以帮助相关人员对发电机的装机容量形式、地点和时间以及电网的整体布局与提供相应的参考。

与此同时,由于这种负荷预测工作而做而进行的实现比较长久,因此结果的精确度在实际应用的过程当中不是很高,往往存在着较大的误差,因此在进行工作规划的时候要留下一定的余地。

1.2 短期以及超短期负荷预测所谓的超短期负荷预测指的是未来一个小时、半个小时、甚至是未来十分钟之类的预测。

这些类型的负荷预测之所以有必要进行下去,是由于可以为电网进行计算在线控制,并且对配电网发电厂等系统进行实时调动指令下达的配合;对于短期负荷预测来说,是在一年之内以月度为单位和以周天小时为单位的负荷预测。

电力系统外文翻译

电力系统外文翻译

外文资料(一)Current density according to the economic section of the wire researchCurrent density according to the economic section at the wire, according to the following formula :jn A '=c nI J (1-1) Jn=max jn I A '(1-2) where Ic------ design sought by the current calculation, unit A; b----- line with the cost of wire cross-section of relations coefficient;β------ rates Potential for the yuan / (kw • h);τ------ maximum load factor, unit of h;α------ cost factors, according to state regulations, can be found on the manual;Figure 1 wire running costs and the annual cross-section curv ejn A ' standards section is not, By the plan (1), we can see that the curve F so there jn A ' corresponding to the lowest point, because the power loss charges section A with the decrease of the reasons, if not envisaged curve Fs play, Section increasewouldinevitably lead to the increase in operating costs. So admission standards section should not only satisfy the minimum requirements of the power loss, but also reduce running costs less because(1) In general, the establishment of factories or load a development process, the initial value is smaller than the design, gradually in order to achieve the expected A'network is completed by the load considered, This is not consistent value, butjnwith the actual situation, in other words, power loss is not designed so much to the imagination;(2) Design calculations indicate the actual load Ic than big design value;(3) F curve relatively flat bottom.A'smaller than Therefore, the selection criteria section, it should be by choicejnthe cross section, as F curve flat bottom, operating costs of less impact, taking into account the load values, as well as changes in the law, Theoretical calculation of the power loss will be larger than the actual value. with the options to save much of the initial investment and the consumption of non-ferrous metals.In the factory power supply system design using Jn wire cross section, Energy losses are still high volume of large factories into line and the electric network in the short occasions application. Method used Jn wire cross section still in use, but it should be noted that this method has the following problems :(1) 1.2 formula of the b value is not constant, the domestic tariff beta value is not uniform, Operating costs of Europium value in different countries should have the period of change.(2) This method is only from the operating expenses for at least the premise, not the investment, operating costs and the overall efficiency.Therefore, the proposed foreign books "at least expenditure," the wire cross section. Under Ic can elect to meet the heating requirements of the specifications 2-3 lead, their investment costs and operating costs are different. High investment costs of cross-section wire resistance by small and less power loss costs, it will be able to choose one of the best programs. But because the wire cross section Size is notcontinuous, but a broken line, in order to solve the lowest value to be used on the dogleg approximation method for the mathematical model, which is relatively more complicated, it has not been applied to engineering practice.(二)Grounding the researchCircuits are grounded in order to prevent high voltages from building up on the conductors, while equipment grounding aims at preventing enclosures from reaching voltages above ground. Grounding thus improves system protection and reliability and provides safety to people standing by.Grounding every circuit, however, makes the system susceptible to excessive currents should a short circuit develop between a live conductor and ground. Thus, not all neutrals of wye-connected loads (especially large motors) should be grounded. Grounding should then be practiced selectively, especially on the primary distribution system, as shown in Fig. -1. In part (a), disconnection of motors M1 and M3 for maintenance of repair deprives the 2400-volt system of a ground. It is preferable toground the system at the source, that is, at the transformer neutral in Fig.-1 (b).2400V13.8kVM1M2M3M4(a)2400V13.8kVM1M2M3M4(b)Fig.2 Circuit grounding done selectively(a) at a few motor neutrals (load);(b) at the transformer neutral (source)Metal enclosures,raceways,and fixed equipments are normally grounded. However,motor and generators well insulated from ground,and metal enclosurs used to protect cables or equipments from physical damage,may be left ungrounded.Aslo,portable tools and home appliances,such as refrigerators and air conditions,need not be grounded if constructed with double insulation.Some ac circuits are required to be ungrounded as,for instance,in anesthesizing locations in hospital.In fact,line isolation monitors are installed in such cases,capable of sounding warning signals.High-voltage services (>1000V) are not necessarily grounded, but they must be so if they supply portable equipment.Metal underground water pipes are normally used for grounding, If their length is judged inadequate, they may be complemented by other means, such as a building metal frame or some underground pipe of tank.中文译文(一)按照经济电流密度选择导线截面的研究按照经济电流密度选择导线截面时,可根据下式:jn A '=c nI J (1-1) Jn=max jn I A '(1-2) 式中 Ic------设计时求得的计算电流,单位为A ;b-----线路造价与导线截面间的关系系数; β------电价,电位为元/(kw·h)τ------最大负荷损耗系数,单位为h ;α------费用系数,根据国家规定,可在有关手册中查到;图1 导线截面与年运行费的关系曲线jn A '未必是标准截面,那么,由图 1可以看出,曲线F 所以出现对应于jn A '的最低点,是因为电能损耗费随截面A 的增大而减小的缘故,设想如果没有曲线Fs 起作用,截面的增加必然引起运行费用的增加。

电气专业毕业设计外文翻译--基于人工智能的长期电力负荷预测

电气专业毕业设计外文翻译--基于人工智能的长期电力负荷预测

附录A: 外文文献及译文第一部分:原文Artificial intelligence in long term electric load forecastingK. Metaxiotis, A. Kagiannas, D. Askounis, J. PsarrasAbstract: Intelligent solutions, based on artificial intelligence (AI) technologies, to solve complicated practicalproblems in various sectors are becoming more and more widespread nowadays. AI-based systems arebeing developed and deployedworldwide in myriad applications, mainly because of their symbolic reasoning,flexibility and explanation capabilities.This paper provides an overview for the researcher of AI technologies, as well as their current use in thefield of long term electric load forecasting (LTELF). The history of AI in LTELF is outlined, leading to adiscussion of the various approaches as well as the current research directions. The paper concludes bysharing thoughts and estimations on AI future prospects in this area. This review reveals that although stillregarded as a novel methodology, AI technologies are shown to have matured to the point of offering realpractical benefits in many of their applications.Keywords: Artificial intelligence; Electric load forecasting; Energy1. IntroductionIn the past two decades. AI has been defined as the study of how to make computers do thingsthat, at the moment, people do better. AI provides powerful and flexible means for obtaining solutions to a variety ofproblems that often cannot be solved by other, more traditional and orthodox methods.This review bears witness to the application of AI technologies in the field of long term electric loadforecasting (LTELF). Certainly, this is not the first paper to review the application of AI basedsystems in energy related problems with varying success. In general, AI developments in the fieldof energy have been reviewed by several authors from various points of view.Taylor and Lubkemanreviewed the applications of knowledge based programming to powerengineering problems, describing prototype projectsdeveloped at North Carolina State University,while the survey of Zhang et al. concerned the use of ES technology in electric powersystems. Ypsilantis and Yee presented a review of ESs for SCADA based power applicationsand Lubarskii et al. discussed the use of ESs for power networks. Since that time, several othersurvey papers have been written invariousenergy related areas.However, this paper has a different focus. Writing a fully comprehensive survey of AI applicationsin energy systems is objectively impracticable. For this reason, our paper aims to create a large knowledgebase for the researcher, introducing him/her to the specific area of AI applications in LTELF andindicating other fields fertile for research.2. AI applications in long term electric load forecasting2.1Expert systemsESsare one of the most commercially successful branches of AI. Welbankdefines an ES asfollows:An expert system is a program, which has a wide base of knowledge in a restricted domain,and uses complex inferential reasoning to perform tasks, which ahuman expert could do.In other words, an ES is a computer system containing a well organised body of knowledge,which emulates expert problem solving skills in a bounded domain ofexpertise. The systemis able to achieve expert levels of problem solving performance, which would normally beachieved by a skilled human, when confronted with significant problems in the domain.The first works in ES application in LTELF were implemented by Rahman and Bhatnagarand Jabbour et al. The objective of these approaches was to use the knowledge, experienceand analytical thinking of experimental system operators. Park et al. made a further step by usingfuzzy logic in an ES for a LTELF problem. In 1990, Ho et al. presented the use of aknowledge based ES in long term load forecasting of a Taiwan power system, while in 1993,Rahman and Hazim tried to generalize his first work. Markovic and Fraissler proposedan ES approach (based on Prolog) for long term load forecasting by plausibility checking ofannounced demand.In 1995, Kim et al. implemented a long term load forecaster by using ANNs and afuzzy ES, while later, Mori and Kobayashi presented an optimal fuzzy inference approach forthe LTELF problem. Ranaweera et al. proposed a fuzzy logic ES model for the LTELFproblem, which used fuzzy rules to incorporate historical weather and load data. These fuzzy ruleswere obtained from historical data using a learning type algorithm.A back propagation neural network with the output provided by a rule based ESwas designedby Chiu et al. for the LTELF problem. To demonstrate that the inclusion of the predictionfrom a rule based ES of a power system would improve the predictive capability of the network,load forecasting was performed on the Taiwan power system. The evaluation of the systemshowed that the inclusion of the rule based ES prediction significantly improved the neural network’s prediction of power load.2.2Artificial neural networksANNs are an information processing technique based on the way biological nervous systems,such as the brain, process information. The fundamental concept of ANNs is the structure of theinformation processing system. Composed of a large number of highly interconnected processingunits (“neurons”) connected into networks, a neural network system uses the human-like techniqueof learning by example to resolve problems. Every neuron applies an input, activationand an output function to its net input to calculate itsoutput. The neural network is configured for a specific application, such as data classification orpattern recognition, through a learning process called “training”.The first researchers who introduced the ANN application in LTELF were Lee et al., whoproposed an innovative ANN methodology for the LTELF problem. Park et al. proposed theuse of a multilayer network with three layers, i.e. one input, one hidden and one output. Thetraining of the network was performed through a simpleback-propagation algorithm. Using loadand weather information, the system produced three different forecast variables. Lee et al. treated electric load demands as a non-stationary timeseries, and they modeled the load profile by a recurrent neural network.In 1992,Peng et al. presented a search procedure for selecting the training cases for ANNsto recognize the relationship between weather changes and load shape, while Ho et al. implementeda multilayer neural network with an adaptive learningalgorithm.Chen et al. proposed an ANN for weather sensitive long term load forecasting, while analternative technique using a recurrent high order neural network was considered by Kariniotakiset al. Papalexopoulos et al. proposed the inclusion of additional input variables, suchas a seasonal factor and a cooling/heating degree into a single neural network.Czernichow et al. used a fully connected recurrent network for load forecasting in whichthe learning database consisted of 70,000 patterns with a high degree of diversity. Mandal et al.applied neural networks for LTELF in which the inputs consisted of the past load data only,and no weather variables were used, while Sforna and Proverbio investigated the application of ANNs in LTELF, through a research project at ENEL, and confirmed their positivecontribution.In1997, Kiartzis et al. presented the Bayesian combined predictor, aprobabilisticallymotivated predictor for LTELF based on the combination of an ANN predictor and two linearregression predictors. The method was applied to LTELF for the Greek Public Power Corporationdispatching center of the island of Crete. Ramanathan et al. made several comparisonsof statistical, time series and ANN methods for the LTELF.In 1998, Sforna reported the implementation of a software tool, called NEUFOR, based onANN technology and specifically designed to meet the operational needs of utility power systemdispatchers regarding online operation, while Papadakis et al. continued to improve theirprevious work. The same goes for Drezga and Rahman. The development of improved neuralnetwork based LTELF models for the power system of the Greek island of Crete, as well as radialbasis function networks and fuzzy neural type networks, were proposed and discussed by Kodogiannisand Anagnostakis in 1999. In the years 2000 and 2001, several researchers dealt withthe application of ANN to the LTELF problem, with varying success.3. ConclusionsElectricity long term load forecasting is important for the power industry, especially in thederegulated electricity market. Proper demand forecasts help the market participants to maximizetheir profits and/or reduce their possible losses by preparing an appropriate bidding strategy.Traditional statistics based linear regression methods need modification to capture the more andmore non-linearities in demand signals under the market conditions.What emerges from this discussion is that AI based systems are becoming more and morecommon decision making tools in LTELF. AI methods for forecasting have shown an ability togive better performance in dealing with the non-linearities andother difficulties in modeling thetime series. The ESs as well as the ANNs have been found to be the most popular for this field.The advantage of these technologies over statistical models lies in their ability to model a multivariate problem without making complex dependency assumptions among input variables.Furthermore, the ANN extracts the implicit non-linear relationship among input variables bylearning from training data.Concluding, we can say that AI techniques, like all other approximation techniques, haverelative advantages and disadvantages. There are no rules as to when a particular technique ismore or less suitable for LTELF. Based on the survey presented here, it is believed that AI offersan alternative “philosophy" which should not be underestimated at all.第二部分:译文基于人工智能的长期电力负荷预测K. Metaxiotis, A. Kagiannas, D. Askounis, J. Psarras摘要:基于人工智能( AI )技术的智能解决方案,由于是为了解决复杂的实际问题,所以在社会各界得到越来越广泛的重视。

(完整word版)电力系统负荷预测及方法(外文翻译)

(完整word版)电力系统负荷预测及方法(外文翻译)

Power system load forecasting methods and characteristics of Abstract: The load forecasting in power system planning and operation play an important role, with obvious economic benefits, in essence, the electricity load forecasting market demand forecast. In this paper, a systematic description and analysis of a variety of load forecasting methods and characteristics and that good load forecasting for power system has become an important means of modern management.Keywords: power system load forecasting electricity market construction Planning1.IntroductionLoad forecasting demand for electricity from a known starting to consider the political, economic, climate and other related factors, the future demand for electricity to make predictions. Load forecast includes two aspects: on the future demand (power) projections and future electricity consumption (energy) forecast. Electricity demand projections decision generation, transmission and distribution system, the sic of new Capacity; power generating equipment determine the type of prediction (.such as peaking units, base load units, etc}.Load forecasting purposes is to provide load conditions and the level of development, while identifying the various supply areas, each year planning for the power consumption for maximum power load and the load of planning the overall level of development of each plan year to determine the load composition.2. load forecasting methods and characteristics of2.1 Unit Consumption ActOutput of products in accordance with national arrangements, planning and electricity intensity value to determine electricity demand. Sub-Unit Consumption Act; Product Unit Consumption; and the value of Unit Consumption Act; two. The projection of load before the key is to determine the appropriate value of the product unit consumption or unit consumption. Judging from China's actual situation, the general rule is the product unit consumption increased year by year, the output value unit consumption is declining. Unit consumption method advantages arc: The method is simple, short-torn load forecasting effective. Disadvantages arc: need to do a lot of painstaking research work, more general, it is difficult to reflect modern economic, political and climate conditions.2.2 Trend extrapolationWhen the power load in accordance with time-varying present same kind of upward or downward trend, and no obvious seasonal fluctuations, but also to find a suitable function curve to reflect this change in trend, you can use the time t as independent variables, timing value of y for the dependent variable to establish the trend model y = f (t). When the reason to believe that this trend will extend to the future, we assigned the value of the variable t need to, you can get the corresponding tune series of the future value of the moment. This is the trend extrapolation.Application of the trend extrapolation method has two assumptions: (1) assuming there is no step Change in load; (2)assume that the development of load factors also determine the future development of load and its condition is unchanged or changed little. Select the appropriate trend model is the application of the trend extrapolation an important part of pattern recognition method and finite difference method is to select the trend model arc two basic ways.A linear trend extrapolation forecasting method, the logarithmic trend forecasting method, quadratic curve trend forecasting method, exponential curve trend forecasting method, growth curve of the trend prediction method. Trend extrapolation method's advantages arc: only need to historical data, the amount of data required for less. The disadvantage is that: If a change in load will cause large errors.2.3 Elastic Coefficient MethodElasticity coefficient is the average growth rate of electricity consumption to GDP ratio of between, according to the gross domestic product growth rate of coefficient of elasticity to be planning with the end of the total electricity consumption. Modules of elasticity law is determined on power development from a macro with the relative speed of national economic development, which is a measure of national economic development and an important parameter in electricity demand. The advantages of this method arc: The method is simple, easy to calculate. Disadvantages arc: need to do a lot of detailed research work.2.4 Regression Analysis MethodRegression estimate is based on past history of load data, build up a mathematical analysis of the mathematical model. Of mathematical statistics regression analysis of the variables in statistical analysis of observational data in order to achieve load to predict the future. Regression model with a linear regression, multiple linear regression, nonlinear regression and other regression prediction models. Among them, linear regression for the medium-torn toad forecast. Advantages arc: a higher prediction accuracy for the medium and the use of short-term forecasts. The disadvantage is that: (1) planning level it is difficult years of industrial and agricultural output statistics; (2) regression analysis can only be measured out the level of development of an integrated electricity load can not be measured out the power supply area of the loading level of development, thus can notbe the specific grid construction plan.2.5 Time Series AnalysisThe load is on the basis of historical data, trying to build a mathematical model, using this mathematical model to describe the power load on the one hand this random variable of statistical regularity of the change process; the other hand, the mathematical model based on the re-establishment of the mathematical expression of load forecasting type, to predict the future load. Time series are mainly autoregressive AR (p), moving average MA (q) and self-regression and n3oving average ARMA (p, q) and so on. The advantages of these methods arc: the historical data required for less, work less. The disadvantage is that: There is no change in load factor to consider, only dedicated to the data fitting, the lack of regularity of treatment is only applicable to relatively uniform changes in the short-term load forecasting situation.2.6 Gray model methodGray prediction is a kind of a system containing uncertain factors to predict approach. Gray system theory based on the gray forecasting techniques may be limited circumstances in the data to identify the role of law within a certain period, the establishment of load forecasting models. Is divided into ordinary gray system model and optimization model for two kinds of gray.Ordinary gray prediction model is an exponential growth model, when the electric load in strict accordance with exponentially growing, this method has high accuracy and required less sample data to calculate simple and testable etc.; drawback is that for a change in volatility The power load, the prediction error largo, does not meet actual needs. And the gray modeloptimization can have ups and downs of the original data sequence transformed into increased exponentially increasing regularity changes in sequence, greatly improving prediction accuracy and the gray model method of application. Gray Model Law applies to short-torn load forecast. Gray predicted advantages: smaller load data requirements, without regard to the distribution of laws and do not take into account trends, computing convenient, short-term forecasts of high precision, easy to test. Drawbacks: First, when the data the greater the degree of dispersion, namely, the greater the gray level data, prediction accuracy is worse; 2 is not very suitable for the long-term power system to push a number of years after the forecast.2.7 Delphi MethodThe Delphi method is based on the special knowledge of direct experience, research problems of judgment, a method for prediction of, also called experts investigation. Delphi method has feedback, anonymity and statistical characteristics. Delphi method advantage is:(1) can accelerate prediction speed and save prediction Cost; (2)can get different but valuable ideas and opinions; (3)suitable for long-term forecasts in historical data, insufficient or unpredictable factors is particularly applicable more. Detect is: (1)the load forecasting far points area may not reliable; (2)the expert opinions sometimes may not complete or impractical.2.8 Expert System ApproachExpert system prediction is stored in the database over the past tow years, even decades, the Hourly load and weather data analysis, which brings together experienced staff knowledge load forecasting, extract the relevant rules, according to certain rules, load prediction. Practice has proved that accurate load forecasting requires not only high-tech support, but also need to reconcile the experience and wisdom of mankind itself: Therefore, you need expert systems such technologies. Expert systems approach is a non-quantifiable human experience translated into a better way But experts systems analysis itself is a time-consuming process, and some complex factors (such as weather factors), even though aware of its load impact, ht}t to accurately and quantitatively determine their influence on the load area is also very difficult. Expert system for forecasting method suitable for medium and long-term load forecast. The advantages of this method: (1)can bring together multiple expert knowledge and experience to maximize the ability of experts; (2) possession of data, information and mort factors to consider a more comprehensive and beneficial to arrive at mart accurate conclusions. The disadvantage is that: (1)do not have the self-learning ability, subject to the knowledge stored in the database limits the total; (2) pairs of unexpected incidents and poor adaptability to changing conditions 2.9 Neural Network MethodNeural network (ANN, Artificial Neural Network) forecasting techniques to mimic the human brain to do intelligent processing, a large number of non-structural. non-deterministic laws of adaptive function. ANN used in short-term load forecasting and long-term load forecast than that applied to be mart appropriate. Because short-term load changes can be regarded as a stationary random process. And long-term load forecasting may be due to political, economic and other major fuming point leading to a mathematical model-based damage. Advantages arc:(1) to mimic the human brain, intelligence processing; (2}a large number of non-structural. non-adaptive function of the accuracy of the law; (3)with the information memory, self-learning, knowledge, reasoning and optimization of computing features. The disadvantage is that:(1) the determination of the initial value can not take advantage of existing system information, easily trapped in local minimum of the state; (2) neural network learning process is usually slow, pooradaptability to sudden events.2.10 Optimum Combination Forecasting MethodOptimal combination has two meanings: First, several forecasting methods from the results obtained by selecting the appropriate a0cight in the weighted average; 2 refers to the comparison of several prediction methods, choose the best or the degree of preparation and the standard deviation of the smallest prediction model forecast. For the combined forecasting method must also noted that the combined forecast is a single forecasting model can not completely correct to describe the changes of the amount predicted to play a role. One can fully reflect the actual law of development of the model predictions agree well with the combination forecasting method than predicted good results. This method has the advantage: To optimize the combination of a wide range of information on a single prediction model, consider the impact of information is also mart comprehensive, so it can effectively improve the prediction. The disadvantage is that: (1) the weight is difficult to determine; (2) all possible factors that play a role in the future, all included in the model, to a certain extent, limit the prediction accuracy improved.2.11 Wavelet analysis and forecasting techniquesWavelet analysis is a time-domain-frequency domain analysis method, it is in the time domain and frequency domain at the same time has good localization properties, and can automatically adjust according to the signal sampling frequency of high and low density, it is cast' to capture and analysis of weak signals and signal, images of any small parts. The advantage is: Can the different frequency components gradually refined using a sampling step, which can be gathered in any of the details of the signal, especially for singular signal is very sensitive to the treatment well or mutation weak signals, their goal is to a signal information into wavelet coefficients, which can easily be dealt with, storage, transmission, analysis or for the reconstruction of the original signal. These advantages determine the wavelet analyses can be effectively applied to load forecasting issues.3. ConclusionLoad forecasting is the electric power system scheduling, real-time control, operation plan and development planning, the premise is a grid dispatching departments and planning departments must have the basic information. Improve load forecasting technology level, be helpful for program management, reasonable arrangement of the electricity grid operation mode for the maintenance plan and the crew, to section coal, fuel-efficient and reduce generating cost, be helpful for formulate rational power construction planning of the power system, improve the economic benefit and social benefit. Therefore, the load forecast has become a power system management modernization realization of the important content.电力系统负荷预测及方法摘要:负荷预测在电力系统规划和运行方面发挥的重要作用,具有明显的经济效益,负荷预测实质上是对电力市场需求的预测。

负荷预测(翻译)

负荷预测(翻译)

IEEE TRANSACTIONS ON INDUSTRY APPLICA TIONS, VOL. 45, NO. 4, JULY/AUGUST 2009多区域负荷预测系统 与大的地理区域舒凡,成员,kittipong methaprayoon ,成员,李伟仁,研究员,IEEE (电气与电子工程师协会)摘要—在电力系统覆盖的地理区域,一个单一的负荷预测模型的整个地区,有时不能保证较好的预测精度。

其中的一个主要的原因是由于负载的多样性,通常造成的天气的多样性,在整个地区。

多区域负荷预测将是一个可行的和有效的产生更多的解决方案准确的预测结果,以及提供区域预报水电费。

然而,一个主要的挑战是如何以最佳方式分割/合并的区域,根据区域负荷天气条件。

本文研究了电力需求从中西部电力气象数据,美国对数据的分析,我们证明了存在的天气和负载的多样性,其控制区域内的再发展基于支持短期的多区域负荷预测系统日前的经营和市场的向量回归。

所提出了多区域预测系统可以发现最佳区域分区在不同的天气和负载条件下,最终实现更准确的预测汇总系统负载。

该预测系统已通过实际测试从系统的数据。

得到不同的结果区域划分方案,验证了所提出的有效性多区域预测系统。

在详细讨论预测的结果也被提出。

指数条款—分散负荷,负荷预测,多区,支持向量回归(SVR )。

一、引言 负荷预测是电力系统运行的一个关键问题教育,规划,营销[ 1 ]。

许多经营决策,如发电调度计划,可靠性分析,并对机组检修计划,基于负荷预测。

未来的新一代规划植物和传输的增强也依赖于负载预测。

因此,负荷预测一直是一个热门的话题电力工业。

到目前为止,各种各样的技术已经被提出预测电力负荷[ 2 ]。

一般来说,大部分的工作集中在预测模型本身,并没有具体注意到已经支付给一个系统的负荷预测大的地理区域,虽然有些问题需要要在实际操作解决。

在电力系统中的大的地理区域,天气和电力需求的多样性在整个地区是一个关键的问题出在影响预测精度。

电力负荷预测方法

电力负荷预测方法

电力负荷预测方法浅析摘要:电力负荷预测是电力系统调度、用电、计划、规划等管理部门的重要工作之一。

本文系统地叙述了几种常用负荷预测方法的实质特点、实用场合及其负荷预测误差度的检验,为电力负荷预测提供参考依据。

关键词:电力负荷预测预测误差组合预测abstract: power load forecasting is one of the most important tasks of the dispatching, electricity, plan, planning departments of power system. this paper describes essential characteristics, applications and load forecast errors of several commonly used load forecasting’s methods, and provides reference for power load forecastingkey words:power load forecasting prediction error combination forecasting中图分类号:f407.61 文献标识码:a 文章编号:2095-2104(2012)负荷预测是电力系统运行管理与建设发展的基础工作,也是长期以来的热点问题.目前国内外研究的方法主要包括基于回归分析、神经网络、灰色理论和最优组合预测方法等[1~3],基本上都属于参数统计法的范畴,因变量对自变量有较强的依赖关系;当假设函数模型成立时,预测精度较高,当假设函数不成立时,预测模型的拟合情况和预测精度都是不理想的;后者降低了自变量对因变量的限制,有较大的适应性,但也可能会失去历史资料所提供的信息,降低模型的解释能力.电力负荷预测是电力系统调度、用电、计划、规划等管理部门的重要工作之一。

电力系统潮流计算软件设计外文原文及中文翻译

电力系统潮流计算软件设计外文原文及中文翻译

电力系统潮流计算软件设计外文原文及中文翻译外文原文及中文翻译Modelling and Analysis of Electric Power SystemsPower Flow Analysis Fault AnalysisPower Systems Dynamics and StabilityPrefaceIn the lectures three main topics are covered,i.e.Power flow an analysisFault current calculationsPower systems dynamics and stabilityIn Part I of these notes the two first items are covered,while Part II givesAn introduction to dynamics and stability in power systems. In appendices brief overviews of phase-shifting transformers and power system protections are given.The notes start with a derivation and discussion of the models of the most common power system components to be used in the power flow analysis.A derivation of the power ?ow equations based on physical considerations is then given.The resulting non-linear equations are for realistic power systems of very large dimension and they have to be solved numerically.The most commonly used techniques for solving these equations are reviewed.The role of power flow analysis in power system planning,operation,and analysis is discussed.The next topic covered in these lecture notes is fault current calculations in power systems.A systematic approach to calculate fault currents in meshed,large power systems will be derived.The needed models will be given and the assumptions made when formulating these models discussed.It will be demonstrated thatalgebraic models can be used to calculate the dimensioning fault currents in a power system,and the mathematical analysis has similarities with the power ?ow analysis,soitis natural to put these two items in Part I of the notes.In Part II the dynamic behaviour of the power system during and after disturbances(faults) will be studied.The concept of power system stability isde?ned,and different types of pow er system in stabilities are discussed.While the phenomena in Part I could be studied by algebraic equations,the description of the power system dynamics requires models based on differential equations.These lecture notes provide only a basic introduction to the topics above.To facilitate for readers who want to get a deeper knowledge of and insight into these problems,bibliographies are given in the text.Part IStatic Analysis1 IntroductionThis chapter gives a motivation why an algebraic model can be used to de scribe the power system in steady state.It is also motivated why an algebraic approach can be used to calculate fault currents in a power system.A power system is predominantly in steady state operation or in a state that could with sufficient accuracy be regarded as steady state.In a power system there are always small load changes,switching actions,and other transients occurring so that in a strict mathematical sense most of the variables are varying with thetime.However,these variations are most of the time so small that an algebraic,i.e.not time varying model of the power systemis justified.A short circuit in a power system is clearly not a steady state condition.Such an event can start a variety of different dynamic phenomena in the system,and to study these dynamic models are needed.However,when it comes to calculate the fault current sin the system,steady state(static) model swith appropriate parameter values can be used.A fault current consists of two components,a transient part,and a steady state part,but since the transient part can be estimated from the steady state one,fault current analysis is commonly restricted to the calculation of the steady state fault currents.1.1 Power Flow AnalysisIt is of utmost importance to be able to calculate the voltages and currents that different parts of the power system are exposed to.This is essential not only in order to design the different power system components such asgenerators,lines,transformers,shunt elements,etc.so that these can withstand the stresses they are exposed to during steady state operation without any risk of damages.Furthermore,for an economical operation of the system the losses should be kept at a low value taking various constraint into account,and the risk that the system enters into unstable modes of operation must be supervised.In order to do this in a satisfactory way the state of the system,i.e.all(complex) voltages of all nodes in the system,must be known.With these known,all currents,and hence all active and reactive power flows can be calculated,and other relevant quantities can be calculated in the system.Generally the power ?ow,or load ?ow,problem is formulated as a nonlinear set of equationsf (x, u, p)=0(1.1)wheref is an n-dimensional(non-linear)functionx is an n-dimensional vector containing the state variables,or states,ascomponents.These are the unknown voltage magnitudes and voltage angles of nodes in the systemu is a vector with(known) control outputs,e.g.voltages at generators with voltage controlp is a vector with the parameters of the network components,e.g.line reactances and resistancesThe power flow problem consists in formulating the equations f in eq.(1.1) and then solving these with respect to x.This will be the subject dealt with in the first part of these lectures.A necessary condition for eq.(1.1) to have a physically meaningful solution is that f and x have the same dimension,i.e.that we have the same number of unknowns as equations.But in the general case there is no unique solution,and there are also cases when no solution exists.If the states x are known,all other system quantities of interest can be calculated from these and the known quantities,i.e. u and p.System quantities of interest are active and reactive power flows through lines and transformers,reactive power generation from synchronous machines,active and reactive power consumption by voltage dependent loads, etc.As mentioned above,the functions f are non-linear,which makes the equations harder to solve.For the solution of the equations,the linearizationy X Xf ?= (1.2)is quite often used and solved.These equations give also very useful information about the system.The Jacobian matrix Xf ?? whose elements are given by j iij X f X f ??=??)((1.3)can be used form any useful computations,and it is an important indicator of the system conditions.This will also be elaborate on.1.2 Fault Current AnalysisIn the lectures Elektrische Energiesysteme it was studied how to calculate fault currents,e.g.short circuit currents,for simple systems.This analysis will now be extended to deal with realistic systems including several generators,lines,loads,and other system components.Generators(synchronous machines) are important system components when calculating fault currents and their model will be elaborated on and discussed.1.3 LiteratureThe material presented in these lectures constitutes only an introduction to thesubject.Further studies can be recommended in the following text books:1. Power Systems Analysis,second edition,by Artur R.Bergen and VijayVittal.(Prentice Hall Inc.,2000,ISBN0-13-691990-1,619pages)2. Computational Methods for Large Sparse Power Systems,An object oriented approach,by S.A.Soma,S.A.Khaparde,Shubba Pandit(Kluwer Academic Publishers, 2002, ISBN0-7923-7591-2, 333pages)2 Net work ModelsIn this chapter models of the most common net work elements suitable for power flow analysis are derived.These models will be used in the subsequent chapters when formulating the power flow problem.All analysis in the engineering sciences starts with the formulation of appropriate models.A model,and in power system analysis we almost invariably then mean a mathematical model,is a set of equations or relations,which appropriately describes the interactions between different quantities in the time frame studied and with the desired accuracy of a physical or engineered component or system.Hence,depending on the purpose of the analysis different models of the same physical system or components might be valid.It is recalled that the general model of a transmission line was given by the telegraph equation,which is a partial differential equation, and by assuming stationary sinusoidal conditions the long line equations, ordinary differential equations,were obtained.By solving these equations and restricting the interest to the conditions at the ends of the lines,the lumped-circuit line models (π-models) were obtained,which is an algebraic model.This gives us three different models each valid for different purposes.In principle,the complete telegraph equations could be used when studying the steady state conditions at the network nodes.The solution would then include the initial switching transients along the lines,and the steady state solution would then be the solution after the transients have decayed. However, such a solution would contain a lot more information than wanted and,furthermore,it would require a lot of computational effort.An algebraic formulation with the lumped-circuit line model would give the same result with a much simpler model ata lower computational cost.In the above example it is quite obvious which model is the appropriate one,but in many engineering studies these lection of the“correct”model is often the most difficult part of the study.It is good engineering practice to use as simple models as possible, but of course not too simple.If too complicated models are used, the analysis and computations would be unnecessarily cumbersome.Furthermore,generally more complicated models need more parameters for their definition,and to get reliable values of these requires often extensive work.i i+diu+du C ’dx G ’dxR ’dx L ’dx u dxFigure2.1. Equivalent circuit of a line element of length dx In the subsequent sections algebraic models of the most common power system components suitable for power flow calculations will be derived.If not explicitly stated,symmetrical three-phase conditions are assumed in the following.2.1 Lines and CablesThe equ ivalent π-model of a transmission line section was derived in the lectures Elektrische Energie System, 35-505.The general distributed model is characterized by the series parametersR′=series resistance/km per phase(?/km)X′=series reactance/km per phase(?/km)and the shunt parametersB′=shunt susceptance/km per phase(siemens/km)G′=shunt conductance/km per phase(siemens/km )As depicted in Figure2.1.The parameters above are specific for the line or cable configuration and are dependent onconductors and geometrical arrangements.From the circuit in Figure2.1the telegraph equation is derived,and from this the lumped-circuit line model for symmetrical steady state conditions,Figure2.2.This model is frequently referred to as the π-model,and it is characterized by the parameters)(Ω=+=impedance series jX R km km km Z )(siemens admittance shuntjB G Y sh km sh km sh km =+= I mk Z km y sh km y sh mkI kmkmFigure2.2. Lumped-circuit model(π-model)of a transmission line betweennodes k and m.Note. In the following most analysis will be made in the p.u.system.Forimpedances and admittances,capital letters indicate that the quantity is expressed in ohms or siemens,and lower case letters that they are expressed in p.u.Note.In these lecture notes complex quantities are not explicitly marked asunder lined.This means that instead of writing km Z we will write km Z when this quantity is complex. However,it should be clear from the context if a quantity is real or complex.Furthermore,we will not always use specific type settings for vectors.Quite often vectors will be denoted by bold face type setting,but not always.It should also be clear from the context if a quantity is a vector or a scalar.When formulating the net work equations the nodeadmittance matrix will be used and the series admittance of the line model is neededkm km 1-km km jb g z y +== (2.1)With22km r g km km kmx r +=(2.2)and 22km x -b km km kmx r += (2.3)For actual transmission lines the series reactance km x and the series resistance km r are both positive,and consequently km g is positive and km b is negative.The shunt susceptance sh y km and the shunt conductance sh g km are both positive for real line sections.In many cases the value of sh g km is so small that it could be neglected.The complex currents km I and mk I in Figure2.2 can be expressed as functions of the complex voltages at the branch terminal nodes k and m:k sh km m k km km E y E E y I +-=)( (2.4)m k m mk )(E y E E y I sh km km +-=(2.5)Where the complex voltages arek j k k e θU E = (2.6)k j k k e θU E =(2.7) This can also be written in matrix form as))(()(m k sh km km km km sh km km mk km E E y y y -y -y y I I ++=(2.8) As seen the matrix on the right hand side of eq.(2.8)is symmetric and thediagonal elements are equal.This reflects that the lines andcables are symmetrical elements.2.2 TransformersWe will start with a simplified model of a transformer where we neglect the magnetizing current and the no-load losses .In this case the transformer can be modelled by an ideal transformer with turns ratio km t in series with a series impedance km z which represents resistive(load-dependent)losses and the leakage reactance,see Figure2.3.Depending on if km t is real ornon-real(complex)the transformer is in-phase or phase-shifting.p k mU m ej θm I km I mkU kej θk U p e j θp Z km 1:t km p k mU m ej θm I km I mkU kej θk U p e j θp Z km t km :1Figure2.3. Transformer model with complex ratio kmj km km e a t ?=(km -j 1-km km e a t ?=) mp k U m ej θm I km I mk U kej θk U p e j θp Z km a km :1Figure2.4. In-phase transformer model 2.2.1In-Phase TransformersFigure2.4shows an in-phase transformer model indicating the voltage at the internal –non-physical –node p.In this model the ideal voltage magnitude ratio(turns ratio)iskm k p(2.9) Since θk = θp ,this is also the ratio between the complex voltages at nodes k and p, km j k j p k pa e U e U E E k p ==θθ(2.10)There are no power losses(neither active nor reactive)in the idealtransformer(the k-p part of the model),which yields0I E I E *mk p *km k =+(2.11) Then applying eqs.(2.9)and(2.10)giveskm mk km mk km -a I I -I I ==(2.12)A B Ck m I mk I kmFigure2.5. Equivalent π-model for in-phase transformerwhich means that the complex currents km I and mk I are out of phase by 180since km a ∈ R.Figure2.5 represents the equivalent π-model for thein-phase transformer in Figure2.4.Parameters A, B,and C of this model can be obtained by identifying the coefficients of the expressions for the complex currents km I and mk I associated with the models of Figures2.4 and 2.5.Figure2.4 givesm km km k km 2km p m km km km E y a E y a E -E y -a I )()()(+==(2.13)m km k km km p m km mk E y E y a -E -E y I )()()(+== (2.14)or in matrix form ))(()(m k km km km km km km2km mk km E E y y a -y a -y a I I =As seen the matrix on the right hand side of eq.(2.15) is symmetric,but thediagonal elements are not equal when 1a 2km ≠.Figure2.5 provides now the following:m k km E A -E A -I )()(+=(2.16)m k mk E C A E A -I )()(++=(2.17)or in matrix form))(()(m k mk km E E C A A -A -B A I I ++= (2.18)Identifying the matrix elements from the matrices in eqs.(2.15) and (2.18) yieldskm km y a A = (2.19)km km km y 1-a a B )(= (2.20)km km )y a -(1C =(2.21) 2.2.2 Phase-Shifting TransformersPhase-shifting transformers,such as the one represented in Figure2.6,are used to control active power flows;the control variable is the phase angle and the controlled quantity can be,among other possibilities,the active power flow in the branch where the shifter is placed.In Appendix A the physical design of phase-shifting transformer is described. A phase-shifting transformer affects both the phase and magnitude of the complex voltages k E and p E ,without changing their ratio,i.e., km j km km k p e a t E E ?== (2.22)Thus, km k p ?θθ+=and k km p U a U =,using eqs. (2.11) and (2.22)km j -km *km mkkm e -a -t I I ?==I km m U m ej θm I mk pkU k ej θk Z km 1:a kme j φkmkm k p ?θθ+=k km p U a U = Figure2.6. Phase-shifting transformer with km j km km e a t ?=As with in-phase transformers,the complex currentskm I and mk I can be expressed in terms of complex voltages at the phase-shifting transformer terminals:m km *km k km 2km p m km *km km E y t -E y a E -E y -t I )()()(+== (2.24)m km k km km p m km mk E y E y t -E -E y I )()()(+==(2.25)Or in matrix form))(()(m k km km km km *km km 2km mk km E E y y t -y t -y a I I =(2.26) As seen this matrix is not symmetric if km t is non-real,and the diagonal matrixelements are not equal if 1a 2km ≠.There is no way to determine parameters A, B,and Cof the equivalent π-model from these equations,since the coefficient km *km y t - ofEm in eq.(2.24)differs from km km y t -in eq.(2.25),as long as there is non zero phase shift,i.e. km t ?R.A phase-shifting transformer can thus not be represented by a π-model.2.2.3Unified Branch ModelThe expressions for the complex currents km I and mk I for both transformersand shifters derived above depend on the side where the tap is located;i.e., they are not symmetrical.It is how ever possible to develop unified complex expressions which can be used for lines,transformers,and phase-shifters, regardless of the side on which the tap is located(or even in the case when there are taps on both sides of thedevice).Consider initially the model in Figure2.8 in which shunt elements have beentemporarily ignored and km j km km e a t ?= and m k j mk mk e a t ?=。

外文翻译--电力负荷预测方法:决策工具

外文翻译--电力负荷预测方法:决策工具

附录3Electric load forecasting methods: Tools for decisionmakingHeiko Hahn, Silja Meyer-Nieberg *, Stefan PicklFakult für Informatik, Universitat der Bundeswehr, 85577 Neubiberg, Germany AbstractFor decision makers in the electricity sector, the decision process is complex with several different levelsthat have to be taken into consideration. These comprise for instance the planning of facilities and anoptimal day-to-day operation of the power plant. These decisions address widely differenttime-horizonsand aspects of the system. For accomplishing these tasks load forecasts are very important. Therefore,finding an appropriate approach and model is at core of the decision process. Due to the deregulationof energy markets, load forecasting has gained even more importance. In this article, we give an overviewover the various models and methods used to predict future load demands.2009 Elsevier B.V. All rights reserved.1. Load forecasts in deregulated marketsDecision making in the energy sector has to be based on accurateforecasts of the load demand. Therefore, load forecasts areimportant tools in the energy sector. Forecasts of different timehorizonsand different accuracy are needed for the operation ofplants and of the complex power syste m itself: The ‘‘system responsefollows closely the load requirement” (Kyriakides and Polycarpou, 2007, p. 392). The decision maker is faced with a multitudeof decision problems on different time-scales as well as on differenthierarchies of the power system: These problems comprisefor instance the determination of an optimal secure scheduling ofunit commitment and energy allocation. But decisions do not havemade only with respect to the day-to-day operation of the powersystem but also with respect to investment decisions on new facilitiesbased on the anticipation of future energy demands. For bothends, reliable forecasts are needed. The deregulation of energymarkets has increased the need for accurate forecasts even more(see e.g. Feinberg and Genethliou, 2005; Kyriakides and Polycarpou,2007). To participate in the market, a player needs an accurateestimate howmuch energy is needed at a certain time. On theone hand, an underestimation of the energy demand by a suppliermay lead to high operational costs because the additional demandhas to be met by procuring energy in the market. An overestimationon the other hand wastes scarce resources (see e.g. Tzafestasand Tzafestas, 2001; Feinberg and Genethliou, 2005; Kyriakidesand Polycarpou, 2007). Furthermore, demand is one of the mainfactors for pricing. LoadDue to the highimportance of accurate load forecasting, the history of this fieldis quite long: A 1987 survey paper (Gross and Galiana, 1987) listsan impressive number of publications devoted to load analysisand forecasting – reaching back as far as 1966 (Heinemann et al.,1966). Up to now, various approaches have been introduced. Theycan be grouped into two main classes: Models and methods whichfollow a more classical approach, i.e., which apply concepts stemmingfrom time series and regression analysis and methods whichbelong to the fields of Artificial and Computational Intelligence.This paper gives a short survey over models and methods forload forecasting. Further survey and review papers are for exampleKyriakides and Polycarpou (2007), Feinberg and Genethliou (2005),Tzafestas and Tzafestas (2001) and Hippert et al. (2001).2. Short-term, medium-term and long-term forecastsAs we have seen, forecasts are made for various purposes: theday-to-day operation of the power system (e.g. Kyriakides andPolycarpou, 2007) requires the prediction of the load for a dayahead whereas the decision whether to undertake major structuralinvestments requires a far longer prediction horizon. Forecasts canbe distinguished therefore firstly by the time-horizon or the leadtime: short-term load forecasts (STLF) usually aim to predict theload up to one-week ahead (Kyriakides and Polycarpou, 2007).Frequently, the term very short-term load forecast is used forforecasts with a time-horizon of less than 24 hours (see Yang,2006, p. 7). Up to now, the main focus in load forecasting has beenon STLF since it is an important tool in the day-to-day operation ofutility systems (see e.g. Gonzalez-Romera et al., 2006). Morerecently with the deregulation of energy markets, more and moreattention is also paid to load forecasts with a greater time-horizon,i.e., medium-term load forecasts. As stated in (Gonzalez-Romeraet al., 2006), medium-term load forecasts enables companies toestimate the load demand for a longer time interval which helpsthem for example in the negotiation of contracts with other companies.Medium-term load forecasts (MTLF) are from one weekto one year. Forecasts aiming at load prediction for more than ayear ahead are usually termed long-term load forecasts (LTLF)(see e.g. Feinberg and Genethliou, 2005). As stated in Kyriakidesand Polycarpou (2007) the time-horizon in LTLF is usually 20 yearsalthough longer lead times of 25–30 years can be found. The differencesin lead times have consequences for the models and methodsapplied and for the input data available and selected. The load demandis influenced by numerous factors – ranging from weatherconditions over seasonal effects to socio-economic factors. Whichinput data has to be selected usually depends on the task and dataat hand. The decision maker, therefore, is not only faced with thetask of selecting an appropriate model type but also with determiningimportant external factors. Both tasks usually depend oneach other. Some general observations can be made, however.conditions whether weather-dependent factors havea significant influence on the prediction. There is a common agreementthat the air temperature is the most important weather influence(see e.g. Hippert et al., 2001; Feinberg and Genethliou, 2005).This was already recognized in the 1930s (Hippert et al., 2001).Generally, the demand is high on cold days which can be attributedto electric heating. Similarly on hot days, the increased usage ofair-conditioning generates a higher demand of energy. In manycountries, this results in aU-shaped and clearly non-linear responsefunction of the load towards the temperature (Hippertet al., 2001). However, the exact shape of the curve depends onthe region, the climatic conditions and of course on the consumers’behavior. Additionally, the designated time-horizon and the availabilityof the datadetermine the input variables. As mentioned in (Tayloret al., 2006) univariate models are standard for very short-termload forecasts for up to 6 hours ahead. Furthermore, it should benoted that sometimes obtaining accurate weather forecasts maybe difficult. Therefore, univariate models are also applied for longerlead times (Taylor et al., 2006; Soares and Souza, 2006).In Kyriakides and Polycarpou (2007) three main groups of inputdata forshort-term load forecasts are identified: seasonal inputvariables, weather forecast variables, and historical load data(Kyriakides and Polycarpou, 2007). Short-term load forecasts usuallyaim at providing the daily, hourly, or half-hourly load and thepeak load (day, week) (see e.g. Tzafestas and Tzafestas, 2001)although even smaller time intervals occur. Forecasting the loadprofile, i.e., the load of the next 24 hours, is also a main target(Tzafestas and Tzafestas, 2001; Hippert et al., 2001).Medium-term load forecasts usually incorporate several additionalinfluences - especially demographic and economic factors.These forecasts often provide the daily peak and average load,although hourly loads are also sometimes given, e.g. Bruhns et al.(2005). In the case of long-term load forecasts, even more indicatorsfor the demographic and economic development have to be takeninto account (Kyriakides and Polycarpou, 2007). These are forinstance thepopulation growth and the gross domestic product.Long-term load forecasting usually aims at predicting the annualload and the peak load (Kyriakides and Polycarpou, 2007).The time series of the loads itself has generally three seasonalcycles: anintra-daily cycle (the daily load curve or the load profile),a weekly cycle, and a yearly seasonal cycle. The weekly cycle usuallyshows two main groups:week-days and weekends. Due toindustrial demand, the load tends to be higher during week-days.The weekend tends to influence the neighboring days so that Mondaysand Fridays are often treated separately. Saturday is also often found to show a different load profile than Sunday. However, theexact weekly pattern depends on the particular region under considerationand furthermore on the season (Hippert et al., 2001).Additionally ‘‘regular” exceptional cases can beTheseexceptional days depend on the calendar date which is also commonlyconsidered as a very important information. Several approaches,however, neglect these exceptional cases (e.g. Hippertet al., 2005; Taylor and McSharry, in press). One of the reasonsfor this is that there are typically only few data available whichcould be used to estimate the model parameters for these day -types. Concerning the seasonal data patterns, different approaches arefollowed.Generally, local and global approaches can be distinguished.Local approaches apply distinct models for each identifiedfeature: for instance a model for each distinct day-type or a hour.However, this may lead towards problems if the data basis is notsufficiently large (Hippert et al., 2001). Therefore, in severalapproachesa single monolithic model is build and the informationis encoded in the input data.3. Models and methodsThis section gives an overview over various approaches for loadforecasting. Many of them are developed for STLF although MTLFhas gained importance which is especially due to the deregulationof electricity markets. The dominance of STLF-methods is also reflectedin the survey papers (Hippert et al.,2001;Tzafestas andTzafestas, 2001; Feinberg and Genethliou, 2005; Kyriakides andPolycarpou, 2007). Only Feinberg and Genethliou (2005) includedan overview over MTLF and LTLF-methods whereas the remaindersfocused on short-term load forecasts. Our own non-exhaustive surveyof over 100 papers shows a similar composition.There are various approaches applied to load forecasting(Kyriakides andPolycarpou, 2007; Feinberg and Genethliou,2005; Taylor and McSharry, in press; Hippert et al., 2001; Tzafestasand Tzafestas, 2001). These range fromregression-based approachesover time-series approaches towards artificial neural networksand expert systems. In the following, a short overview oversome of the models and methods is given. The error measure mostfrequently used to assess the performance of a model – at least inthe literature found – is the mean absolute percentage error(MAPE) defined by 1100Tt t t MAPE y y T ==-∑with t y the real value atpoint t and ty the forecast. 3.1. Classical time series and regression methodsStatistical approaches require an explicit mathematical modelwhich gives the relationship between load and several input factors.Several classical models are applied for load forecasting, forexample regression-based methods, time series methods, statespace models and Kalman-filtering. In the following, we focus on regression-based and time series models.3.1.1. Regression-based modelsRegression models are quite common in load forecasting(Kyriakides andexternal factors, for instanceweather and calendar information or customer types (Feinbergand Genethliou, 2005). Mainly, linear regression is used – theinfluence of the temperature is usually modeled non-linearly,however.Regression methods are relatively easy to implement.A further advantage is that the relationship between input andoutput variables is easy to comprehend. Regression models alsoallow relatively easy performance assessments (see e.g. Bruhnset al., 2005). As reported in Kyriakides and Polycarpou (2007),there may be inherentproblems in identifying the correct model,though, which are due to the complex non-linear relationship betweenthe load and the influencing factors. Further drawbacks arereported in Kyriakides and Polycarpou (2007). In thefollowing,two examples are presented. Many more approaches appear inthe literature – for example regression models based on localpolynomial regression for STLF (Zivanovic, 2001), non-parametricregression (Charytoniuk et al.,1998), or robust regression methods(Jin et al., 2004). A description of furtherregression-based approachescan be found in Kyriakides and Polycarpou (2007) andFeinberg and Genethliou (2005).Hor et al. developed a multiple regression model and analyzedthe impact of weather variables on the load demand for Englandand Wales (Hor et al., 2005). They used data from 1989 to 1995for model training and from 1996 to 2003 for testing the accuracy.Their aim was to provide an accurate model for a long-term predictionof the monthly demand. The regression model was basedon two types of input variables: weather-dependent factors, i.e.,temperature, wind speed W V , rainfall r M , relative humidity Hr ,and hours of sunshine Ms andsocio-economic factors, i.e., theGross Domestic Product. Further socio-economic factors, for instancethe population growth, were eliminated. First of all, theyanalyzed the relationship between load and temperature andfound anon-linear dependence. In order to cope with this non-linearityseveral derived variables: heating degree days (HDD), coolingdegree days (CDD), and enthalpy latent days (ELD) wereintroduced. The last variable takes the influence of humidity onair-conditioning into account. Three linear regression modelswere finally proposed: model 1 includes the CDD, HDD, and theELD0123456.A w s rE CDD HDD ELD V M M ααααααα=++++++(1) Model 2 substitutes these values with the raw temperature and relativehumidity while model 3 differs by model 1 only by a substitutionof the ELD with therelative humidity. All models were thenadjusted to take socio-economic factors into account. Model 1 and3 were found to perform best with mean absolute percentage errors(MAPE) of 1.98% and 2.1% – although model 2 is alsorelatively goodwith a MAPE of 2.69%.Bruhns et al. (2005) presenteda non-linear regression model forMTLF. Their model is devoted to an hourly prediction of the load.The load was decomposed into a weather-dependent i Phc andweather-independent part i Pc ,i i i i i Gaussian (Bruhnset al., 2005). The weather-independent part coverestrends,seasonalbehavior, and calendar effects. The weather-dependent partis assumed to be mainly influenced by the temperature and thecloud cover. The relationship between temperature and load is fittedby a non-linear model differentiatingbetween a heating part(temperature above a certain threshold) and a cooling part (temperaturebelow a certain threshold) of the weather sensitive partof the load. The temperature does not enter the equations directly.First of all, it is exponentially smoothed to reflect the inertia tovariations inbuildings (Bruhns et al., 2005). Afterwards, it is averagedagain with the observed temperature and – in the case of theheating part – with the cloud cover. Thisreflects the assumptionthat there is a mixed response to the temperature observed andto the temperature felt inside buildings (Bruhns et al., 2005). Thisaggregated temperature then enters the threshold function.Fortheweather-independent part of the load a product of two Fourierseries was used. The first depended on the hour, the second on theday-type. The authors do not report the MAPE but the scale-dependentrootmean squared error (RMSE). However, they note that thepredecessor of their model which applied one Fourier seriesdepending on the hour already achieved a MAPE of 2% for yearahead forecasts (known weather) and 1.5% for day ahead forecasts(Bruhns et al., 2005).3.1.2. Time-series approachesTime-series approaches (Box and Jenkins, 1970) are among theoldest methods applied in load forecasting. They can be distinguishedon several levels. First of all, there are univariate (Tayloret al., 2006; Taylor and McSharry, in press ) and multivariatemethods. The former are usually used for very short-termloadforecasts whereas the latter are applied for all time-horizons.The time series is usually assumed to be linear. It should be noted,though, that the assumption of linearity usually does not comprisethe influence of the temperature. In this case, there appearsto be an unanimous agreement that the non-linearity of thisrelationshiphas to be preserved in the model (Moral-Carcedo andVicens-Otero, 2005). Several authors apply non-linear models(Hor et al., 2006).A very simple class is the so-called autoregressive movingaverage or ARMA models(Brockwell and Davis, 1991). In short, astationary process ()0t t X ≥is called anARMA(p,q)-process if 11p qt k t k t m t q k m X X Z Z φθ--==-=+∑∑ (2)holds. The process ()t Z must be white noise with zero mean andconstantstandarddeviation σThe main steps consist of determiningthe order p of theautoregressive (AR)-part and q of the movingaverage (MA)-part and ofestimating the coefficients (Brockwell andDavis, 1991). Common approaches for the estimation of the coefficientsuse Maximum-Likelihood or variants of leastnoise. However, in Huang and Shih (2003) an ARMA-modeling procedurefor STLF was presented which also allows for non-Gaussiannoise. In the following, let B denote the lag or backshift operator, i.e.,1.t t BX X -=Setting ()11p k k k B B φ-Φ=-∑ and ()11q m m m B B θ-Θ=+∑,(2) can bewritten shortly as ()()t t B X B Z Φ=Θ.In Huang et al. (2005) an ARMAX-model ()()()()t t B L t B Z B U Φ=Θ+ψ with ()1km m m B B ψ=ψ=∑and exogenous variable t U forone-day and one-week ahead hourly load forecasts for four exampleseasons in 1998 was presented. Theexogenous variable in thecase denoted the temperature. Instead of the classical algorithmsto determine the model and to estimate the coefficients, particleswarm optimization (PSO) (Eberhart and Kennedy, 1995) was appliedto determine both the order of the model and the coefficients.Particle swarm optimization is a so-called swarm-based optimizationmethod and belongs to the class ofComputational Intelligence(CI) methods.The ARMA-model can be extended to so-called autoregressiveintegrated moving average (ARIMA) models. This model types usesdifferencing in order to cope with non-stationarity. An ARIMA(p,d,q)-model is thus()()()1d t t B B X B Z Φ-=Θ. The parameterd is required to be an integer. Themodels above are univariate linearmodels most often used for short-term load forecasts. As mentionedin Taylor et al. (2006), univariate ARIMA-models are oftenused as sophisticated benchmark models in STLF.Amjady used a modified ARIMA-model to predict the hourlyload demand and the daily peak loads in STFL (Amjady, 2001).His modified ARIMA-models takes not only past loads but alsoestimates of past loads into account. The estimates were providedby experienced human experts. Thus, in a sense the modelincorporates human knowledge. First of all, Amjady identifiedfourdifferent day-types in the load data of the Iranian nationalgrid: Saturday, Sunday to Wednesday, Thursday and Friday andpublic holidays. These models were separated again into modelsfor hot days (average temperature above 23 degrees) and colddays. In total, 16 models were used. The parameters wereestimatedusing data from 1996 to 1997 from the Iranian nationalgrid and tested on data for 1998. The MAPE of the models rangedfrom 1.48% (Sunday to Wednesday, hot) to 1.99% (public holidays,cold).The ARIMA/ARMA-models can be extended to take seasonalityinto account. Several models have been developed for this task.For example, in Espinoza et al. (2005) a periodic autoregressivemodel (PAR) was used of order p,11,,t s s t s p t p s t y C y y φφε--=++++ (3)(see Espinoza et al., 2005) to represent the periodic dynamics of theseries. The parameters are allowed to vary across the NS seasons. In Espinoza et al. (2005) aload curves whereas the monthlyand weekly seasonality was coveredbyintroducing dummyvariables.Taylor et al. (2006) and Taylor and McSharry (in press) presentedacomparison of several univariate methods for short-termload forecasting. They analyzed four main models and two benchmarkfunctions. The models chosen were double seasonal ARIMA(with-in day and with-in week cycle), exponential smoothing fordouble seasonality, artificial neural network, and a regressionmethod with principal component analysis (PCA) (Taylor et al.,2006). The exponential smoothing method adapted by the authorsis an extension of the classical seasonal Holt-Winter smoothingmethod to incorporate two seasonal cycles()()12111t t t t t s t S y S S T D W αα----=+-+ (4) ()()111t t t t T S S T γγ--=-+- (5)()121t t t s t t s y D D S W δδ--=+- (6) ()211t t t s t t s y W w W S D ω--=+- (7) ()()()()()121121t t k t t t t s k t s s t t s t s y k S kT D W y S T D W φ--+-+---=++-+ (8)(see Taylor et al., 2006). The variables t s and t T denote thesmoothed level and trend;t D and tW are seasonal indices (intradayand intra-week) and ()t y k is the forecast at t k +from the startingpoint t (Taylor et al., 2006). The greek parameters (except φ) arethe smoothing parameters to be determined. The parameter φ is anadjustment for first-order autocorrelation.The comparison was based on two time series: the hourly demandfor Rio de Janeiro in 1996 (30 weeks, 5th May 1996–30th November 1996) and thehalf-hourly demand for England andWales in 2000 (30 weeks, 27th March –22nd October). Several errormeasures were considered, however, only the MAPE was reportedsince the other error measures lead to similar results.The exponential smoothing method emerged as the bestapproach.In Taylor and McSharry (in press) the study was extended toan evaluation of six univariate methods based on the electricitydemand of ten European countries. Again, the double seasonalexponential smoothing method was found to lead to the best results – with a MAPE below 2% even for the longest lead time of24 hours. As it can be seen, time series based approaches are verycommon. However, several drawbacks are reported. As regression-based approachestime-series approaches may suffer fromnumerical instabilities (Huang and Shih,3.2. Artificial intelligence and computational intelligence methodsComputational intelligence is a relatively new research field.The expression computational intelligence is commonly used to referto the fields of fuzzysystems, artificial neural networks (ANN),evolutionary computation, and swarm intelligence. Of these fields,neural networks are the subtype which is most often applied inload forecasting.3.2.1. Neural networksNeural networks are modeled after the basic working principleof human brains. They consist of several neurons. A neuron receivesinformation over its inputnodes and aggregates the information.Afterwards, it determines its activation and propagatesits response over the output node to other neurons. Neuralnetworksare very frequently applied for load forecasting (see e.g.Hippertet al. (2001) for a survey). As stated in Hippert et al. (2005), in1998 a software based on neural networks technology was used byover 30 US electric utilities. Several subtypes of neural networksexist (see e.g. Bishop, 1995). In load forecasting, for example, radialbasis function networks (Ranaweera et al., 1995;Gonzalez-Romeraet al., 2006), self-organizing maps (Becalli et al., 2004) for clusteringand recurrent neural networks (Senjyu et al., 2004; Tran et al.,2006) are used. However, feed-forward neural networks (or multilayerperceptron) are the subtype which is most often applied(Hippert et al., 2001, 2005;Gonzalez-Romera et al., 2006; Becalliet al., 2004; Ringwood et al., 2001). A feed-forward network consistsof several successive layers of neurons with one input layer,several hidden layers, and an output layer. The neuronsareconnectedusing weight vectors and neither feedback norinterlayerconnectionsexist. A neuron i thus takes the output of its kinputneurons, computes theweighted sum, subtracts a so-calledbias hi and applies the activation function a (), i.e.,()1ni ik k i k y a w x θ==-∑.The basic learning or weight-adjustingprocedureis back-propagation (a form of steepest descent) which propagatesthe error backwards and adjusts the weightsaccordingly(Bishop, 1995). Frequently, only one hidden layer is used (seeforinstance Becalli et al. (2004), Fidalgo et al. (2007) and Hippertet al. (2005)). Hippert et al. (2005) provided a comparisonof large neural networks (neural networks with a large numberof neurons and weights) with several classical approaches. Theclassical approaches ranged from naive forecasting methods oversmoothing filters and combination of smoothing filters with linearregression. Furthermore, hybrids of smoothing filters andneural networks were considered. The task was to forecast the24 hours load profile based on data from a local utility inRio de Janeiro. Used for building, testing, and validating the forecastmodel were the hourly loads and thetemperature from April1996 to December 1997. Hippert et al. (2005) found large neuralnetworks to perform best – not only with the smallest MAPE(2.35–2.65%) but also with acan be seen as competitivewith other models as far as the forecasting of load profiles isconcerned.4. ConclusionsLoad forecasting is very important for decision processes in theelectricity sector. In this paper, we have given an overview overcommonly used input variables and forecast targets. Afterwards,we presented several classes of models and methods. Several recenttrends can be identified: Support vector regression hasemerged as a relatively new and competitive method for load forecasting.Furthermore, more and more attention is focused on hybridapproaches.Load forecasting is not only important to provide accurate estimatesfor the operating of the power system but also as a basis forenergy transactions and decision making in energy markets. Theaccuracy of forecasts is a very crucial factor: A decision maker inthe energy sector has the need of accurate forecasts since mostof the decisions are necessarily based on forecasts of future demands.One of the first decisions to be made is therefore the selectionof an appropriate model. This depends on the problem and thesituation currently under consideration. Therefore, no general recommendationscan be given.AcknowledgementThe authors wish to thank especially the anonymous reviewer #1 for hishelpful comments附录4电力负荷预测方法:决策工具摘要对于电力部门的决策者来说,决策过程是复杂的,必须考虑不同的标准。

  1. 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. 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. 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)。

电力系统负荷预测及方法(外文翻译)Power system load forecasting methods and characteristics of Abstract: The load forecasting in power system planning and operation play an important role, with obvious economic benefits, in essence, the electricity load forecasting market demand forecast. In this paper, a systematic description and analysis of a variety of load forecasting methods and characteristics and that good load forecasting for power system has become an important means of modern management.Keywords: power system load forecasting electricity market construction Planning1.IntroductionLoad forecasting demand for electricity from a known starting to consider the political, economic, climate and other related factors, the future demand for electricity to make predictions. Load forecast includes two aspects: on the future demand (power) projections and future electricity consumption (energy) forecast.Electricity demand projections decision generation, transmission and distribution system, the sic of new Capacity; power generating equipment determine the type of prediction (.such as peaking units, base load units, etc}.Load forecasting purposes is to provide load conditions and the level of development, while identifying the various supply areas, each year planning for the power consumption for maximum power load and the load of planning the overall level of development of each plan year to determine the load composition.2. load forecasting methods and characteristics of2.1 Unit Consumption ActOutput of products in accordance with national arrangements, planning and electricity intensity value to determine electricity demand. Sub-Unit Consumption Act; Product Unit Consumption; and the value of Unit Consumption Act; two. The projection of load before the key is to determine the appropriate value of the product unit consumption or unitconsumption. Judging from China's actual situation, the general rule is the product unit consumption increased year by year, the output value unit consumption is declining. Unit consumption method advantages arc: The method is simple, short-torn load forecasting effective. Disadvantages arc: need to do a lot of painstaking research work, more general, it is difficult to reflect modern economic, political and climate conditions.2.2 Trend extrapolationWhen the power load in accordance with time-varying present same kind of upward or downward trend, and no obvious seasonal fluctuations, but also to find a suitable function curve to reflect this change in trend, you can use the time t as independent variables, timing value of y for the dependent variable to establish the trend model y = f (t). When the reason to believe that this trend will extend to the future, we assigned the value of the variable t need to, you can get the corresponding tune series of the future value of the moment. This is the trendextrapolation.Application of the trend extrapolation method has two assumptions: (1) assuming there is no step Change in load; (2)assume that the development of load factors also determine the future development of load and its condition is unchanged or changed little. Select the appropriate trend model is the application of the trend extrapolation an important part of pattern recognition method and finite difference method is to select the trend model arc two basic ways.A linear trend extrapolation forecasting method, the logarithmic trend forecasting method, quadratic curve trend forecasting method, exponential curve trend forecasting method, growth curve of the trend prediction method. Trend extrapolation method's advantages arc: only need to historical data, the amount of data required for less. The disadvantage is that: If a change in load will cause large errors.2.3 Elastic Coefficient MethodElasticity coefficient is the average growthrate of electricity consumption to GDP ratio of between, according to the gross domestic product growth rate of coefficient of elasticity to be planning with the end of the total electricity consumption. Modules of elasticity law is determined on power development from a macro with the relative speed of national economic development, which is a measure of national economic development and an important parameter in electricity demand. The advantages of this method arc: The method is simple, easy to calculate. Disadvantages arc: need to do a lot of detailed research work.2.4 Regression Analysis MethodRegression estimate is based on past history of load data, build up a mathematical analysis of the mathematical model. Of mathematical statistics regression analysis of the variables in statistical analysis of observational data in order to achieve load to predict the future. Regression model with a linear regression, multiple linear regression, nonlinear regression and other regressionprediction models. Among them, linear regression for the medium-torn toad forecast. Advantages arc: a higher prediction accuracy for the medium and the use of short-term forecasts. The disadvantage is that: (1) planning level it is difficult years of industrial and agricultural output statistics; (2) regression analysis can only be measured out the level of development of an integrated electricity load can not be measured out the power supply area of the loading level of development, thus can notbe the specific grid construction plan.2.5 Time Series AnalysisThe load is on the basis of historical data, trying to build a mathematical model, using this mathematical model to describe the power load on the one hand this random variable of statistical regularity of the change process; the other hand, the mathematical model based on the re-establishment of the mathematical expression of load forecasting type, to predict the future load. Time series are mainlyautoregressive AR (p), moving average MA (q) and self-regression and n3oving average ARMA (p, q) and so on. The advantages of these methods arc: the historical data required for less, work less. The disadvantage is that: There is no change in load factor to consider, only dedicated to the data fitting, the lack of regularity of treatment is only applicable to relatively uniform changes in the short-term load forecasting situation.2.6 Gray model methodGray prediction is a kind of a system containing uncertain factors to predict approach. Gray system theory based on the gray forecasting techniques may be limited circumstances in the data to identify the role of law within a certain period, the establishment of load forecasting models. Is divided into ordinary gray system model and optimization model for two kinds of gray.Ordinary gray prediction model is an exponential growth model, when the electric load in strict accordance with exponentiallygrowing, this method has high accuracy and required less sample data to calculate simple and testable etc.; drawback is that for a change in volatility The power load, the prediction error largo, does not meet actual needs. And the gray model optimization can have ups and downs of the original data sequence transformed into increased exponentially increasing regularity changes in sequence, greatly improving prediction accuracy and the gray model method of application. Gray Model Law applies to short-torn load forecast. Gray predicted advantages: smaller load data requirements, without regard to the distribution of laws and do not take into account trends, computing convenient, short-term forecasts of high precision, easy to test. Drawbacks: First, when the data the greater the degree of dispersion, namely, the greater the gray level data, prediction accuracy is worse; 2 is not very suitable for the long-term power system to push a number of years after the forecast.2.7 Delphi MethodThe Delphi method is based on the special knowledge of direct experience, research problems of judgment, a method for prediction of, also called experts investigation. Delphi method has feedback, anonymity and statistical characteristics. Delphi method advantage is:(1) can accelerate prediction speed and save prediction Cost; (2)can get different but valuable ideas and opinions; (3)suitable for long-term forecasts in historical data, insufficient or unpredictable factors is particularly applicable more. Detect is: (1)the load forecasting far points area may not reliable;(2)the expert opinions sometimes may not complete or impractical.2.8 Expert System ApproachExpert system prediction is stored in the database over the past tow years, even decades, the Hourly load and weather data analysis, which brings together experienced staff knowledge load forecasting, extract the relevant rules, according to certain rules, load prediction.Practice has proved that accurate load forecasting requires not only high-tech support, but also need to reconcile the experience and wisdom of mankind itself: Therefore, you need expert systems such technologies. Expert systems approach is a non-quantifiable human experience translated into a better way But experts systems analysis itself is a time-consuming process, and some complex factors (such as weather factors), even though aware of its load impact, ht}t to accurately and quantitatively determine their influence on the load area is also very difficult. Expert system for forecasting method suitable for medium and long-term load forecast. The advantages of this method: (1)can bring together multiple expert knowledge and experience to maximize the ability of experts; (2) possession of data, information and mort factors to consider a more comprehensive and beneficial to arrive at mart accurate conclusions. The disadvantage is that: (1)do not have the self-learning ability, subjectto the knowledge stored in the database limits the total; (2) pairs of unexpected incidents and poor adaptability to changing conditions2.9 Neural Network MethodNeural network (ANN, Artificial Neural Network) forecasting techniques to mimic the human brain to do intelligent processing, a large number of non-structural. non-deterministic laws of adaptive function. ANN used in short-term load forecasting and long-term load forecast than that applied to be mart appropriate. Because short-term load changes can be regarded as a stationary random process. And long-term load forecasting may be due to political, economic and other major fuming point leading to a mathematical model-based damage. Advantages arc:(1) to mimic the human brain, intelligence processing; (2}a large number of non-structural. non-adaptive function of the accuracy of the law; (3)with the information memory, self-learning, knowledge, reasoningand optimization of computing features. The disadvantage is that:(1) the determination of the initial value can not take advantage of existing system information, easily trapped in local minimum of the state; (2) neural network learning process is usually slow, poor adaptability to sudden events.2.10 Optimum Combination Forecasting MethodOptimal combination has two meanings: First, several forecasting methods from the results obtained by selecting the appropriate a0cight in the weighted average; 2 refers to the comparison of several prediction methods, choose the best or the degree of preparation and the standard deviation of the smallest prediction model forecast. For the combined forecasting method must also noted that the combined forecast is a single forecasting model can not completely correct to describe the changes of the amount predicted to play a role. One can fully reflect the actual law of development of the model predictions agree well with the combination forecasting method than predictedgood results. This method has the advantage: To optimize the combination of a wide range of information on a single prediction model, consider the impact of information is also mart comprehensive, so it can effectively improve the prediction. The disadvantage is that: (1) the weight is difficult to determine; (2) all possible factors that play a role in the future, all included in the model, to a certain extent, limit the prediction accuracy improved.2.11 Wavelet analysis and forecasting techniquesWavelet analysis is a time-domain-frequency domain analysis method, it is in the time domain and frequency domain at the same time has good localization properties, and can automatically adjust according to the signal sampling frequency of high and low density, it is cast' to capture and analysis of weak signals and signal, images of any small parts. The advantage is: Can the different frequency components gradually refined using a sampling step, which can be gathered in any of the details of the signal, especially for singular signal is very sensitive tothe treatment well or mutation weak signals, their goal is to a signal information into wavelet coefficients, which can easily be dealt with, storage, transmission, analysis or for the reconstruction of the original signal. These advantages determine the wavelet analyses can be effectively applied to load forecasting issues.3. ConclusionLoad forecasting is the electric power system scheduling, real-time control, operation plan and development planning, the premise is a grid dispatching departments and planning departments must have the basic information. Improve load forecasting technology level, be helpful for program management, reasonable arrangement of the electricity grid operation mode for the maintenance plan and the crew, to section coal, fuel-efficient and reduce generating cost, be helpful for formulate rational power construction planning of the power system, improve the economic benefit andsocial benefit. Therefore, the load forecast has become a power system management modernization realization of the important content.电力系统负荷预测及方法摘要:负荷预测在电力系统规划和运行方面发挥的重要作用,具有明显的经济效益,负荷预测实质上是对电力市场需求的预测。

相关文档
最新文档