DOI 10.1017S000000000000000 Printed in the United Kingdom Representation in case-based reas
脉冲电流作用下电子风应力模型与计算示例
d ′N A ，式中 d ′ 和 M 为镁密度和摩尔质量。Mg M 的摩尔质量等于 Mg 的相对原子量， 24.3 。因此， d ′ = 1.736 × 103 kg ⋅ m −3 , M = 24.3 g mol [8] 。 N A 为
首先采用位错动力学计算电子风应力。电子密度 ne = Avogadro常数， 6.02 ×
文章引用: 曹富荣. 脉冲电流作用下电子风应力模型与计算示例[J]. 凝聚态物理学进展, 2017, 6(1): 1-5. https:///10.12677/cmp.2017.61001
曹富荣
子力学的电子风应力与实际应力相差很大，造成量子力学方法计算值偏低的原因是没有考虑合金中存在 的第二相阻碍电子运动的情况。
进行下列国际单位与导出单位换算
= Ω m 2 ⋅ kg ⋅ s −3 ⋅ A −2
C= s ⋅ A
N = m ⋅ kg ⋅ s −2
因此，
ΩCA N = m m2
= Few 4.42 × 103 N m
σ ew =
4 Few l 4 Few 4 4.42 × 103 N m = × = × =70.6 MPa 3π 2.66 × 10−6 m 3π d πd 2
摘
要
电致塑性是利用电场和脉冲电流改善材料塑性的方法。为了揭示脉冲电流作用下的电子风影响塑性的机 理，在Conrad电子风力模型的基础上，提出电子风应力模型。在复杂体系的LAZ922 (Mg-Li-Al-Zn)合金 中给出电子风应力计算实例。计算发现，基于位错动力学的电子风应力与实际应力十分吻合，而基于量
( 0.209 − 3.12 ×10 ) m = 7.26 ×10
−2
s m
12
凝聚态物理类纯计算的文章 杂志
凝聚态物理类纯计算的文章杂志
凝聚态物理是研究物质在固态或液态条件下的性质和行为的学科。
纯计算的文章和杂志在这一领域中起着重要作用，因为它们通过模拟和计算的方式帮助我们理解和预测物质的性质。
在这个领域中，有一些知名的期刊和杂志出版了大量的纯计算文章，这些文章涵盖了各种主题，包括但不限于电子结构、凝聚态系统的动力学、磁性、超导性、拓扑材料等等。
其中一些知名的期刊包括Physical Review Letters，Physical Review B，Journal of Applied Physics，Journal of Chemical Physics等。
这些期刊经常刊登纯计算的文章，这些文章通过密度泛函理论、量子蒙特卡洛方法、分子动力学模拟等计算手段，探索了凝聚态物理中的一些重要问题。
纯计算的文章可以涉及到各种不同的技术和方法，例如第一性原理计算、格林函数方法、紧束缚模型等。
这些方法为研究人员提供了研究凝聚态物理的强大工具，帮助他们理解材料的电子结构、磁性和光学性质等方面。
在这些纯计算文章中，研究人员通常会介绍他们的模拟方法、
理论基础以及他们的计算结果，并与实验数据进行比较。
这样的文
章有助于推动凝聚态物理领域的发展，为新材料的设计和性能优化
提供重要的指导。
总的来说，纯计算的文章在凝聚态物理领域扮演着重要的角色，它们通过模拟和计算帮助我们理解和预测物质的性质，推动着这一
领域的发展。
希望我的回答能够对你有所帮助。
NBTI degradation From physical mechanisms to modelling
Introductory Invited PaperNBTI degradation:From physical mechanisms to modellingV.Huarda,*,M.Denaisb,c,C.ParthasarathybaPhilips R&D Crolles,850rue Jean Monnet,38926Crolles,FrancebSTMicroelectronics,Central R&D Labs,850rue Jean Monnet,38926Crolles,France cLaboratoire Mate´riaux et Microe ´lectronique de Provence (L2MP –UMR CNRS 6137)–ISEM,Maison des Technologies,Place Georges Pompidou,83000Toulon,FranceReceived 18January 2005Available online 26April 2005AbstractAn overview of the evolution of transistor parameters under negative bias temperature instability stress conditions commonly observed in p-MOSFETs in recent technologies is presented.The physical mechanisms of the degradation as well as the different defects involved have been discussed according to a systematic set of experiments with different stress conditions.According to our findings,a physical model is proposed which could be used to more accurately pre-dict the transistor degradation.Finally,based on our new present understanding,a new characterization methodology is proposed,which would open the way to a more accurate determination of parameter shifts and thus allowing imple-menting the degradation into design rules.Ó2005Elsevier Ltd.All rights reserved.1.IntroductionBias temperature instability (BTI)is a degradation phenomenon occurring mainly in MOS Field Effect Transistors (MOSFETs),known since the late 1960s [1,2].Even though the exact root causes of the degrada-tion are not yet well understood,it is now commonly admitted that under a constant gate voltage and an ele-vated temperature a build up of positive charges occurs either at the interface Si/SiO 2or in the oxide layer lead-ing to the reduction of MOSFET performances.Never-theless,this degradation remained marginal for many years,especially when compared to hot carrier injection(HCI)degradation,thanks to the exclusive use of buried channel devices at this time.As a result of aggressive scaling of MOS transistors,the thickness of the gate oxide layer was decreased down to 1.4nm or even lower.In order to improve the transis-tors performances,nitrogen atoms were introduced into the oxide layer by different nitridation processes,but mostly by thermal annealing.This nitridation step is in-tended both to give a better control on the gate leakage current and to avoid the boron atoms,used to dope the polysilicon gate,to flow through the oxide into the sub-strate.Besides,the general trend was that most of the de-vices turned out to be surface-channel devices instead of buried-channel ones in recent technologies to counter the short-channel effects inherent of the downscaling process and to improve the performances.As a conse-quence of both the introduction of the nitridation pro-cess step and the use of surface-channel devices,many0026-2714/$-see front matter Ó2005Elsevier Ltd.All rights reserved.doi:10.1016/j.microrel.2005.02.001*Corresponding author.E-mail address:vincent.huard@ (V.Huard).Microelectronics Reliability 46(2006)1–23researchers ascribed an enhanced BTI-like degradation of p-MOSFETs under negative bias and elevated tem-peratures,the so-called NBTI effect[3–5].This work aims to investigate the evolution of tran-sistor parameters under NBTI stress conditions,leading to a general definition of the Negative Bias Temperature Instability in p-MOSFETs in recent technologies.Sys-tematic sets of experiments have been performed with different stress conditions to identify the physical mech-anisms lying behind the degradation.According to our findings,the interface traps creation is not the sole source of degradation but a major hole trapping effect also occurs.This trapping behaviour is at the origin of a strong recovery effect which makes an accurate mea-surement of the degradation difficult.We propose a new characterisation methodology which gives a more accurate view of the actual degradation.2.Device fabrication and experimental detailsThe devices used in this study were n-and p-MOS-FETs with dual gate process,i.e.n+and p+polysilicongate,respectively,with pure or nitrided(NO)gate oxide layers.Source and drain junctions were formed with ar-senic/phosphorous ion implantations for n-MOSFET and a boron-based implantation for p-MOSFET.Exper-iments were carried out on MOSFETs with oxide thick-ness ranging from 1.6nm to10nm.The MOSFETs have gate lengths ranging from0.1l m to10l m and typ-ically a width of10l m.NBTI stress was applied with the gate electrode held at a low constant negative bias(rang-ing fromÀ0.75V toÀ3.5V)under a temperature ranging from25°C to200°C while the source/drain and n-well electrodes were grounded.In order to characterize the NBTI effect,we primarily used a conventional methodol-ogy based on periodic stops during the stress to measure MOS parameters and/or the two-level charge pumping (CP)current as a measure of the interface traps density. The CP curves are measured with a frequency of105Hz and low gate bias V gl ranging fromÀ0.7V to0.3V and a voltage swing from1V to1.5V.A new characterization methodology will be introduced later in the paper,as it is based on recentfindings.Most of the experiments were carried out at125°C unless it is mentioned differently.2.1.Bias temperature instabilities2.1.1.Degradation of electrical parametersBy definition,bias temperature instabilities are ob-served when either a capacitor or a transistor is stressed at relatively high temperatures(typically ranging from 80°C to150°C)under a low and constant gate voltage while the source/drain and well electrodes are grounded. The symmetry of stress conditions along the channel proves that this degradation is not related to channel carrier transport.Fig.1shows the typical evolution of I dÀV g curves in linear regime for a p-MOSFET once degraded under NBTI stress conditions for10,000s. Generally,it is observed that after an NBTI stress,the saturated drain current value(I dsat)is reduced.The for-ward and reverse values of the saturated drain current both degrade identically,demonstrating the symmetry of the stress.By the same time,the threshold voltage (V th)is increased as well as the S/D series resistance.Fi-nally,the mobility,as described through the g mÀV g curve is also reduced(cf.Fig.1).Altogether,the degra-dation of these parameters demonstrates the build-up of positive charges close or at the interface of Si/SiO2and yields to a lower level of performance for the transistor. For short stress times,the off-leakage current is decreas-ing due to the shift of the I dÀV g curves linked to the threshold voltage shift.Nevertheless,in some cases,an increase of the GIDL observed at lowfields may over-come this decrease and limit the performances of the cir-cuit by an increased consumption.The instabilities exist in most of the configurations, for either p-MOSFETs or n-MOSFETs,and whatever a negative bias and/or a positive bias is applied,except for the n-MOSFET under positive bias,which does ex-hibit almost no degradation.Nevertheless,as shown in Fig.2,applying NBTI stress conditions(i.e.negative gate voltage)on p-MOSFETs represents the most degrading case.That is why the rest of this paper will mainly focus on the mechanisms of p-MOSFET degra-dation under NBTI conditions,but all the conclusions would also apply to the other configurations.2V.Huard et al./Microelectronics Reliability46(2006)1–232.2.Interface traps creation under NBTI stress conditionsSo far,the microscopic details of the NBTI degrada-tion are not clearly understood but there is a general agreement to say that there is generation of traps at the Si–SiO2interface during negative BT aging.2.3.Nature of interface trapsDue to the lattice mismatch between the bulk silicon and the silicon dioxide and considering the amorphous nature of the dielectrics,some Si atoms at the interface are left unbound when the great majority is bound to oxygen atoms(or nitrogen atoms for the case of nitrided oxides).This trivalent Si atom at the Si/SiO2interface has an unpaired valence electron in a dangling orbital (dangling bond),and is often called Pb centers[6].On (100)-oriented wafers,commonly used for ICs,two de-fects named Pb0and Pb1have been detected by electron spin resonance(ESR)methods,and showed to result from strain relaxation at the interface.These two types of defects have slightly different local atomic configura-tions and so are expected to have slightly different elec-trical behaviours.In both cases,these defects have an amphoteric nature.This means that the dangling orbital can be occupied by zero,one or two electrons,which would make the same defect positively charged(donor-like),neutral or negatively charged(acceptor-like), depending on the Fermi level at the interface[7,8].Dur-ing the consecutive process steps,these dangling bonds are generally annealed by hydrogen atoms,which create SiH bonds at the interface.Though considerably improving the initial parameters of the transistor,these bonds might be broken during the operating lifetime of the device,which in turn will lead to a degradation of its parameters.2.4.Role of charge transportIn most of wearout mechanisms,such as hot-carrier injection,oxide breakdown or electromigration,experi-ments pointed out that the charge transport activated directly or indirectly the degradation.Naturally, whether or not the charge transport is related to the NBTI degradation is a question,which has to be an-swered.It should be pointed out that the symmetrical nature of the stress with no drain-to-source potential drop implies that there is no charge transport along the channel.The only remaining displacement of chan-nel carriers is based on their thermal activity.Concern-ing the role of carriersflowing through the oxide layer by either direct tunnelling and/or Fowler–Nordheim mechanism,it is interesting to compare both ultra-thin oxides(about2nm-thick)and thicker oxides(about 6.5nm-thick)degradations.In the latter case,the NBTI degradation can be important for these oxides when stressed with a low oxidefield below the detection limit of Fowler–Nordheim current(cf.Fig.3).In that case, the tunnelling probability of carriers through the oxide layer is close to zero which is not the case for thinner oxides under similar oxidefield.Nevertheless,the NBTI degradation monitored in this case is rather similar in spite of the different conduction probabilities through the oxide layer.In agreement with many researchers, we have to conclude that the NBTI degradation isV.Huard et al./Microelectronics Reliability46(2006)1–233closely related to the presence of‘‘cold channel holes’’. This assumption is also supported by the small level of degradation for an n-MOSFET under PBTI conditions (cf.Fig.2),where only negligible hole densities can be found on both sides of the oxide layer.Nevertheless, the exact mechanism of degradation involving cold holes remains unknown at this point.2.5.Influence of channel hole populationIf the channel cold holes are responsible for the NBTI degradation,thefirst thing to check out is the ef-fect of varying the hole population by keeping the other stress conditions constant.This configuration can be ob-tained experimentally by two different ways.Thefirst one is to increase the initial threshold voltage V th0by adding an additional implant.For a similar gate voltage value V g,the oxidefield value would be similar(due to the channel potential grounded to zero)but the channel hole population(proportional to V gÀV th)would de-crease when the threshold voltage is increased.Another way is to change the initial threshold voltage value V th0 by applying a positive bulk bias.In this case,for an increasing positive bulk bias,the threshold voltage value is increased and so that the channel hole population is decreased.Fig.4shows that though V gÀV th is varying by about10%,no changes can be observed in the inter-face traps creation dynamics.In the latter case,some precautions are taken in order to avoid additional degra-dation due to hot holes injection resulting from impact ionisation phenomenon induced by electronsflowing through the oxide layer.Basically,for gate voltage up to2.5V,hot electrons coming from the gate demon-strated no impact on electrical parameters.Under hot carrier stress configuration,degradation rates have been found similar whatever the presence of electronsflowing through the gate.Fig.4shows that,over a relatively large range of channel hole population(large gate volt-ages range for varying bulk voltages),the electrical parameters shifts are similar and not impacted by vary-ing the hole population(i.e.the bulk voltage).As a con-clusion,if the channel holes are responsible for the observed degradation,their population is not a limiting factor.Ourfindings are in agreement with conclusionsof Mitani et al.[9].2.6.Gate voltage or oxidefield dependenceSymmetrically,for different bulk biases,the gate volt-age V g is changed in a way to keep V gÀV th constant and so that the channel hole population is made similar. In this case,only the oxidefield is modified.When the oxidefield is increased,the electrical parameters show a clear increase of their degradation(cf.Fig.5).These results demonstrate the importance of the oxidefield and/or the gate voltage in the degradation.The question of whether the oxidefield or the gate voltage is the driving factor of the degradation remains. Typically,gate voltage dependence is linked to a carrier-energy driven degradation similarly to the case of the gate oxide breakdown and the Channel Hot Carrier (CHC)-induced degradation[10,11].But,as discussed in Fig.3,NBTI degradation occurs also for thick oxides for oxidefields where the gate leakage current is below the detection limit.Therefore,it is difficult to assign the NBTI degradation to energetic carriersflowing through the oxide,either holes or electrons,such as pro-posed in[12].In order to investigate the oxidefield4V.Huard et al./Microelectronics Reliability46(2006)1–23dependence,the interface traps creation under NBTI degradation is investigated for different gate oxide thick-nesses,ranging from2.1nm to10nm-thick,considering both pure oxides and nitrided oxides with various nitrid-ation processes.Fig.6a shows that for pure oxides stressed with a similar gate voltage the interface traps creation is reduced when the oxide thickness is in-creased.Summarizing a set of pure oxides with four dif-ferent thicknesses stressed with various gate voltages for three different stress times,Fig.6b shows that for a given stress time the number of interface traps created are identical for similar oxidefields.These results clearly demonstrate that the oxidefield is the driving force of the interface traps creation during NBTI degradation for pure oxides[13,14].It is questionable if the incorporation of nitrogen atoms into the oxide network will modify this oxide-field dependence.As already discussed previously in the introduction,incorporation of nitrogen is known to en-hance the NBTI degradation.Following the approach that NBTI degradation is solely linked to the creation of interface traps,it means that the nitrogen atoms should modify the properties of the interface.Several possibilities have been already discussed in literature as the introduction of mechanical stresses in the atomic structure close to the SiO2/Si interface,a catalytic role [15],an increase of Pb1proportion with respect to the Pb0center[16],a reduction of the activation energy of SiH bond[17].Whatever the mechanism which might be involved here,the incorporation of nitrogen atoms should modify the interface degradation rate.To solveout that question,both n-MOSFETs and p-MOSFETS devices with either pure or nitrided oxides have been stressed under NBTI stress conditions with thicknesses ranging from 2nm to 10nm.Fig.7shows that the inter-face traps creation is identical for pure and nitrided oxi-des over a wide range of oxide thicknesses.It means that for interface traps creation the oxide field is the major driving force and not the nature of the oxide itself.A crucial point is raised here because the strong impact of the incorporation of nitrogen on the NBTI degrada-tion is well documented in the literature.But,the inter-face traps creation is not modified by the presence of nitrogen atoms.This contradiction shows that the inter-face traps cannot be the sole root cause of the device parameter shifts.This point will be discussed further in this paper.2.7.Temperature dependenceBesides the importance of the oxide field,the NBTI degradation is also activated with temperature.Fig.8shows the influence of the temperature on the interface traps creation for temperatures ranging from 50°C to 200°C under similar oxide field.The time dynamics present two main characteristics:power law behaviour at low stress times and saturation phenomenon for long stress times.Besides it should be noticed that for low stress times,the power law exponent increases with tem-perature.Fig.9shows that it increases linearly with tem-perature.The combination of these two factors implies that the apparent activation energy E a (as determined through degradation levels in an Arrhenius plot)at a given time is not a constant value.Fig.10shows that E a varies with the stress time,increasing for low stress times and finally decreasing for long stress times.As a consequence,the interface traps creation is a non-Arrhenius phenomenon [18],which requires a deeper analysis to understand both the temperature dependence and at the end the physical mechanisms lying behind.This analysis is required in order to be able to determine realistic extrapolation laws for various temperatures.2.8.Time dynamicsUnderstanding the physics that lies behind the NBTI degradation requires to understand not only the oxide field and temperature dependences of the interface traps creation but also the time dynamics.It is also a strong requirement in order to develop an accurateextrapola-6V.Huard et al./Microelectronics Reliability 46(2006)1–23tion model for device lifetime.A characteristic feature of the interface traps creation during NBTI degradation is its fractional power law time dependence(cf.Fig.11). As most of the published data,our experiments span a range from0.2to0.3for the power law exponent.Nev-ertheless,the fractional time dependence of NBTI degra-dation decreases at longer stress times and indicates a tendency towards saturation.Several phenomenological models have been pro-posed to explain the formation of interface traps associ-ated with NBTI degradation.Jeppson et al.[19] proposedfirst a diffusion-controlled mechanism to ex-plain the observed time dependence of interface trap generation.Other authors suggested a similar mecha-nism and examined the time dependence of different charged diffusing species[4,20].Their common assump-tion is that as reaction-limited time dependence obeys a linear relationship,the observed fractional time depen-dence has tofind its origin in a diffusion-limited mode. Diffusion of hydrogenated species away from the inter-face could possibly explain the power law dependence (t a)of the interface traps creation down to t1/2[4],and indeed,at the time,observed values of power law expo-nents ranged between0.75and0.5.However,in more re-cent years,time-dependencies below t1/2and saturation effects are common for sub-micron devices.Recent stud-ies proposed new evolutions of such approach to explain power law exponents ranging about0.25[21,22].In light of the analysis of such Reaction–Diffusion(RD)models, it is possible to distinctly observefive different regimes of evolution(cf.Fig.12from[21]).At short times(t<s reac) (regime1),the system is reaction limited with a charac-teristic slope of1,linked directly to the dissociation en-ergy of SiH bonds.In this approach,it is important to notice that all bonds are identical and the system can be described by a single dissociation energy E d.In re-gime2,the reaction is in equilibrium but theflux of hydrogen away from the interface is negligible.Regime 3is characterized by the hydrogen diffusion limited time dependence described by a power law exponent,which is independent of the oxidefield and/or the temperature but is only determined by the nature of the diffusing hydrogenated species.In regime4,the power law expo-nent increases up to0.5due to hydrogen diffusion into the gate which is supposed to occur with infinite diffu-sion velocity.Finally,in regime5,the generationslowsdown due to saturation of the process and the lack of new bonds to be broken.In order to check the validity of this approach to de-scribe the NBTI degradation,it is important to get some new insights into the interface traps creation during NBTI degradation.One way is to apply preliminary stresses(pre-stress)on the devices in order to break some SiH bonds present at the interface previous to the stress.By doing so,according to the RD model, the regime1should not be impacted since its slope and time constant are only defined by the dissociation energy E d of the bonds.But,a parallel shift downwards of the regime3power law part is expected due to the reduction of the maximum saturation level linked to the maximum number of SiH bonds present previously to the stress.Actually,Fig.13shows that pre-stressing a transistor modifies the reaction-limited part,with a decreasing linear slope(i.e.increasing associated dissoci-ation energy E d)when pre-stress lasts longer.It strictly means that SiH bonds can have various dissociation energies.Concerning the interface traps generation, many authors[23,24]have already reported that the de-fect activation energy of SiH bonds at the interface show a broadened Fermi derivative distribution g(E,r)(with r about0.1eV).In case of distributed dissociation ener-gies,for a given pre-stress time,the slope of the linear part is proportional to the number of SiH bonds for that particular dissociation energy over the characteristic time constant related to this dissociation energy.Our analysis led for two oxidefields on large number of pre-stress times shows that the dissociation energies can befitted by a broadened Fermi derivative distribu-tions with a spread r about0.1eV(cf.Fig.14),as found by other experimental approaches[23,24].In conclusion, we have presented in this section experimental proofs that the SiH bonds at the interface have dispersed disso-ciation energies according to a broadened Fermi deriva-tive distribution.This is an important statement,which will be driven the way we understand and model the interface traps creation.2.9.Model for interface traps creationBreaking a Si–H bond at the interface is often de-scribed by afirst-order reaction such asRðt;sÞ¼1ÀeÀt sÀÁð1Þwhere s represents the time constant of the reaction and is supposed to be directly related to the dissociation en-ergy of the bond E d.Dissociation energy is defined in8V.Huard et al./Microelectronics Reliability46(2006)1–23this case either as the energy to break the SiH bond or the migration barrier the H atoms have to pass over to be released.For degradation times shorter than the time constant(t<s),the degradation rate presents a linear behaviour with time and so a constant defect generation rate P gen=1/s.As shown previously,SiH bonds at the interface do not present a single dissociation energy E d but a contin-uum of energies which,in agreement with other authors [23,24],will be further described by a broadened Fermi derivative distribution g(E d,r)due to disorder-induced variationsgðE d;rÞ¼1re E dmÀE drÀÁ1þe E dmÀE drÀÁ2ð2Þwhere E dm is the median dissociation energy and r is the spread of the distribution with experimental values about0.1eV.In this approach,every single bond is broken accord-ing to afirst-order equation,but each of them with a specific time constant depending on their own dissocia-tion energy.In consequence,given a range of bond ener-gies,lower-energy bonds would be broken relatively quickly leaving higher-energy bonds to be broken more slowly.The degradation rate results from the combina-tion of every single defectfirst-order equation rate, which has been analytically derived as[25]D N it it max ðtÞ¼Z1gðE d;rÞRðt;sðE dÞÞd E d/11þtsÀÁÀað3Þassuming s¼s0expðE dðE oxÞÞand a¼kT for s min<t<s,T being the bond temperature and s min the time constant of the weakest defect.The resulting evolution of the interface traps genera-tion with stress time according to Eq.(3)is shown in Fig.15.For very short times(t<s min),the degradation is lin-ear,with a slope linked to the defect generation rate of the weakest bonds.For longer stress times,more and more different bonds participate to the degradation, yielding to power-law dependence such as described in Eq.(3)with afinal saturation when less and less bonds are left to be broken.Fig.16shows that experimentaldynamics for various temperatures can be well repro-duced by Eq.(3),even the saturation effect that is clearly visible at higher temperatures.An important point in describing the temperature behaviour of the interface traps creation is the capability of the model to predict the temperature dependence of the power law exponent.The linear temperature depen-dence of the power law exponent was already pointed out above in Fig.9.In Fig.17,according to Eq.(3), the spread of the distribution r is determined byfitting experimental power law exponentsÕevolution and is found to be about0.1eV,as previously deduced by non-related methodologies.Due to the various temperature measurements,not only the spread of the distribution is known but it is also possible to have a close idea of what is the saturation value N it max.For a given value of oxidefield and various temperatures,studying the evolution of the time con-stant s,it is possible to extract the constant s o for 2.1nm nitrided oxide.Its value was found to be 1.34·10À8s in this case(cf.Fig.18).This value was found to be identical for various oxidefields.Once thisV.Huard et al./Microelectronics Reliability46(2006)1–239constant is known,it is possible to study the variation of s with oxidefield.This study yields to the determination of the mean dissociation energy E dm,function of theoxidefield.Fig.19shows that the dissociation energy evolves almost linearly with the oxidefield with aflat band limit of about1.5eV,which spans in the range of theoretical values(1.5-1.8eV)proposed by Pantelides et al.[26]for the migration barrier;depending if the hydrogen species migrate away in the substrate or in the oxide.Fig.20shows that using this model allows a good evaluation of the evolution of the apparent activation energy of the interface traps creation.It has to be noted here that if the overall interface traps creation phenom-enon seems to have a non-Arrhenius behaviour,every single bonds considering its own dissociation energy follow an arrhenius behaviour.The deviation from the10V.Huard et al./Microelectronics Reliability46(2006)1–23Arrhenius behaviour only occurs through the existence of the distribution of dissociation energies of the SiH bonds.Besides,this set of parameters allows reproducing not only ultra-thin oxides but also thicker oxides as shown in Fig.21(Nit creation for pure thick oxides with various oxidefields).In conclusion,we have developed a consistent physi-cal-based model of the interface traps creation,which al-lows reproducing all features including oxidefield and temperature dependence over a large range of oxide thickness.3.Threshold voltage degradationSo far,we have made the assumption that the NBTI degradation is only related to the creation of interface traps.We have carefully studied how they are created and how it is influenced by the various stress parameters. But,to understand the impact of the NBTI degradation up to circuit level,it is important to make the link with device parameters such as threshold voltage.As NBTI degradation is mainly a build-up of charges at the inter-face in a symmetrical configuration along the channel, the threshold voltage parameter is more relevant to de-scribe the degradation than other parameters such as the saturated drain current.3.1.Methodology of measurementsFor the NBTI stress characterization,the devices are typically stressed under a constant gate voltage,gener-ally higher than V dd in order to benefit from an acceler-ated degradation,while the source,drain and bulk are grounded.But the stress is periodically interrupted on a linear or a logarithmic time scale,and device parame-ters(threshold voltage,drive current,etc.)are measured at nominal voltage to monitor the degradation.This ap-proach is based on the assumption that the degradation is permanently generated and cannot be removed once the stress is switched off.But,as already shown by Ershov et al[27],inserting a delay between stress inter-ruption and measurements yields to a recovery of at least a part of the degradation.Similar experiments were led measuring both the threshold voltage and the inter-face traps creation through CP measurement.As shown in Fig.22,artificially increased delay between stress interruption and measurements yields also in our case to a partial recovery of the threshold voltage shift.But it is important to notice that the number of interface traps created during the stress is similar.By the way,it is important here to consider the number of interface traps as determined through the integration of the bell-like CP curve and not to make a direct link with the maximum value of the CP current I cpmax.Actually,dur-ing the recovery phase,I cpmax is slightly reduced but the width of the bell-like CP curve is also changed.Leading the integration of this curve shows that the total number of interface traps created during the stress remains con-stant in spite of the reduction of I cpmax.The combination of the unchanged interface traps density,the partial recovery of threshold voltage and the modified edges of the bell-like CP curve with an increased delay points out the fact that the NBTI-induced threshold voltage degradation is not only related to the creation of inter-face traps but a second component has to be taken intoaccount.。
仿生超浸润核桃壳滤料制备及其采油污水处理性能研究
第52卷第7期表面技术2023年7月SURFACE TECHNOLOGY·315·仿生超浸润核桃壳滤料制备及其采油污水处理性能研究时维才1，陈俊旭2,3，宋奇1，陈普2,3，刘平2,3，张友法2,3（1.江苏油田石油工程技术研究院，江苏 扬州，225009；2.东南大学 材料科学与工程学院，南京；3.江苏省先进金属材料高技术研究重点实验室，南京 211189）摘要：目的通过对油田含油污水处理用核桃壳滤料进行润湿性改性，研究其是否能够有效提高核桃壳滤料对采出水的过滤效果，并选出最优改性方法，验证其改性稳定性和长效性。
方法选用粒径为1~3 mm的核桃壳滤料颗粒，经清洗干燥后分别用含氟的超双疏改性液、含辛基硅烷的超疏水超亲油改性液，以及含气相二氧化硅和环氧树脂的超亲水改性液进行润湿性改性，随后对其润湿性分离性能进行评价，并选出最优改性方案，研究改性后滤料的实地服役能力。
结果以核桃壳滤料为基础，通过表面修饰改性方法制备出具有超疏水/超疏油、超疏水/超亲油、超亲水/水下超疏油特性的核桃壳滤料，并通过过滤实验比较了未处理核桃壳滤料和这3种滤料的过滤效果，其过滤效率分别为57.62%、65.77%、61.26%、84.01%，反冲洗液的含油量分别为487.1、172.4、505.8、786.4 mg/L，超亲水/水下超疏油滤料的过滤效率和反洗液含油量最高，表明超亲水/水下超疏油核桃壳滤料可以提升过滤效果，并增强其反洗能力，可作为填充滤料；现场试验结果表明，经超亲水滤料过滤后的水质显著改善，浊度最高降低了83.4%，反冲洗周期≥24 h，展现出良好的应用价值和潜力；进一步探究了核桃壳滤料粒径、滤柱高度、过滤批次等参数对过滤效果的影响，明确指出若要达到较为理想的过滤效果，需对滤料进行级配，且填充高度需合理。
结论超亲水/水下超疏油滤料的过滤效率和反洗液含油量最高，可以提升过滤效果，并增强反洗能力，能够有效改善油田采出水现场过滤表现。
零日病毒传播模型及稳定性分析
用背景，有针对性地引入新的节点状态或调整节点 状态转化机制对病毒传播机理进行研究[13-17]。如针 对高级持续性威胁(Advanced Persistent Threat, APT)攻击和病毒潜伏特性，王刚等人[13]引入了潜 伏状态，提出了易感-潜伏-感染-隔离-移除-易感 (Susceptible–Escape-Infected–Quarantine-Removed–Susceptible, SEIQRS)模型，研究了基于潜 伏隔离机制下的病毒传播规律。文献[11]通过研究 一类新型混合攻击病毒，将该类病毒命名为“去二 存一”病毒并在SEIQRS模型的基础上根据该类病 毒的扩散机理，构建了相应的病毒传播模型。Wang 等人[14]考虑到现有的病毒传播模型由于简化近似， 对大型网络中病毒传播分析造成准确性损失这一情 况，构造了离散时间吸收马尔可夫过程来精确地描 述病毒的传播，并通过仿真分析论证了该方法的准 确性。为了更加精准地刻画病毒在智能校园网上的 传播过程，Wang等人[15]考虑感染病毒个体间的差 异性，将病毒感染个体的进化过程扩展到整个网络 中，建立了智能校园网的病毒传播差分模型从而大 大提高了智能校园网的安全性与鲁棒性。文献[16] 在考虑级联故障普遍存在于复杂网络中的这一情 况，提出了基于局部负荷重分配原则的新型级联失 效模型，建立了SIR病毒传播与级联失效的交互模 型：SIR- c模型，为网络拓扑和路由策略的管理和 优化提供了理论参考。文献[17]考虑实际网络中节 点可以随机移动的情况，基于平均场理论提出了一 个移动环境下网络病毒传播的数学模型，并验证了 这一模型的合理性。这些研究揭示了病毒传播的一 般规律，为病毒的有效防控提供了理论基础。然而 不同病毒传播模型有其适用范围，零日病毒传播机 理相对复杂、隐蔽性强、防御难度大且破坏性大， 需要具体问题具体分析，在现有研究成果的基础 上，结合零日病毒特点研究零日病毒的传播规律及 防控手段。
《基于进化保守性及几何结构相似性的蛋白质与核酸相互作用位点的研究》范文
《基于进化保守性及几何结构相似性的蛋白质与核酸相互作用位点的研究》篇一一、引言随着生物信息学和计算生物学的发展，蛋白质与核酸之间的相互作用研究逐渐成为生命科学研究的重要领域。
蛋白质与核酸的相互作用是细胞内各种生物过程的基础，如基因表达、信号传导等。
理解蛋白质与核酸的相互作用位点对于揭示生命活动的本质具有重要意义。
本文旨在基于进化保守性和几何结构相似性，研究蛋白质与核酸相互作用位点，以期为相关疾病的诊断和治疗提供新的思路。
二、进化保守性在蛋白质与核酸相互作用位点研究中的应用进化保守性是生物分子在进化过程中保持相对稳定的重要特征。
在蛋白质与核酸相互作用的研究中，进化保守性可以用于预测和验证相互作用位点。
通过比较不同物种的基因组序列，可以找出在进化过程中保持不变的氨基酸残基或核苷酸序列，这些序列很可能参与了蛋白质与核酸的相互作用。
在研究方法上，可以利用多序列比对和进化树构建等技术，分析蛋白质与核酸序列的进化关系。
通过比较不同物种的序列，可以确定哪些位点是保守的，从而推测这些位点可能是蛋白质与核酸相互作用的关键位点。
此外，还可以利用结构生物学和分子动力学模拟等技术，进一步验证这些位点的功能和作用机制。
三、几何结构相似性在蛋白质与核酸相互作用研究中的应用几何结构相似性是描述蛋白质与核酸相互作用的重要指标。
通过比较蛋白质与核酸的结构，可以找出它们之间的互补性和匹配性，从而确定相互作用的关键区域。
在研究方法上，可以利用结构生物信息学和分子对接等技术，分析蛋白质与核酸的三维结构。
通过比较它们的结构，可以找出相互作用的界面和关键残基。
此外，还可以利用量子化学计算和分子动力学模拟等技术，进一步探究相互作用的具体机制和动力学过程。
四、综合进化保守性和几何结构相似性研究蛋白质与核酸相互作用位点将进化保守性和几何结构相似性结合起来，可以更准确地研究蛋白质与核酸的相互作用位点。
首先，通过多序列比对和进化树构建等技术，找出进化上保守的氨基酸残基或核苷酸序列。
Characterizing the properties of carbon nanotubes
Characterizing the properties ofcarbon nanotubesCarbon nanotubes (CNTs) have been the subject of extensive research due to their unique structural, electronic, mechanical, and thermal properties. CNTs are cylindrical tubes of carbon atoms, having a diameter of a few nanometers and a length of several micrometers. The walls of CNTs are made of graphene sheets that are rolled up into cylinders, resulting in a seamless tube with a hollow core. The properties of CNTs depend on their diameter, length, chirality, and defects, which can be controlled during the synthesis process.One of the most important properties of CNTs is their high aspect ratio, which is the ratio of their length to diameter. CNTs can have aspect ratios of up to 100,000, which makes them the strongest known materials, with tensile strengths up to 63 GPa. The strength of CNTs comes from their sp2 hybridized carbon bonds, which make the tubes extremely stiff and resilient. CNTs are also highly flexible, and can bend and twist without breaking, enabling them to be used in a wide range of applications.Another important property of CNTs is their electrical conductivity. CNTs are excellent conductors of electricity, with an electrical conductivity of up to 1x107 S/m, which is higher than that of copper. The conductivity of CNTs is dependent on their diameter and chirality, with smaller diameter tubes being more conductive than larger diameter tubes. The high conductivity of CNTs makes them a promising material for electronic and optoelectronic applications, such as transistors, sensors, and solar cells.CNTs also possess exceptional thermal conductivity, which is the ability to conduct heat. CNTs have an extremely high thermal conductivity of up to 3500 W/mK, which is higher than that of any other known material. The high thermal conductivity of CNTs makes them ideal for use in thermal management applications, such as heat sinks and nanocomposites.Furthermore, CNTs are highly hydrophobic, meaning that they repel water. This property makes them useful in applications where water resistance is required, such as in coatings and membranes. CNTs are also resistant to chemical corrosion and oxidation, which makes them highly durable and long-lasting.However, CNTs also have some limitations that need to be addressed. One of the major challenges is their toxicity. While CNTs have shown great promise in medical applications, such as drug delivery and cancer therapy, their potential toxicity to cells and tissues is a cause of concern. Studies have shown that CNTs can cause lung damage and inflammation in rodents, raising questions about their safety for human use. Therefore, it is important to thoroughly evaluate the toxicity of CNTs before using them in biomedical applications.In conclusion, CNTs are a remarkable material with unique and exceptional properties that make them suitable for a wide range of applications. Their high strength, electrical and thermal conductivity, hydrophobicity, and chemical stability make them a promising material in the fields of electronics, energy, and healthcare. However, their potential toxicity needs to be addressed before they can be widely used in biomedical applications. Understanding the properties of CNTs is essential for developing new applications that can exploit their exceptional properties while minimizing their drawbacks.。
具有学习效应的半导体晶圆制造绿色车间调度问题研究
第３０卷 第４期运 筹 与 管 理Ｖｏｌ．３０，Ｎｏ．４２０２１年４月ＯＰＥＲＡＴＩＯＮＳＲＥＳＥＡＲＣＨＡＮＤＭＡＮＡＧＥＭＥＮＴＳＣＩＥＮＣＥＡｐｒ．２０２１收稿日期：２０１９ ０４ １７基金项目：国家自然科学基金资助项目（７１８４０００３）；上海理工大学科技发展资助项目（２０１８ＫＪＦＺ０４３）作者简介：董君（１９８５ ），女，河南焦作人，博士研究生，研究方向：智能算法，生产调度等；叶春明（１９６４ ），男，安徽宣城人，教授，博士生导师，研究方向：工业工程、生产调度等。
具有学习效应的半导体晶圆制造绿色车间调度问题研究董君１，２， 叶春明１（１．上海理工大学管理学院，上海２０００９３；２．河南工学院，河南新乡４５３０００）摘 要：针对最小化最大完工时间、总碳排放以及总拖期时间的具有学习效应的半导体晶圆制造绿色车间调度问题，构建了双影响因素的新型学习效应模型，提出了改进的多元宇宙优化算法，并对其收敛性进行证明。
通过对初始种群进行反向学习、宇宙个体进行莱维飞行扰动和对外部档案中的个体进行邻域搜索变异更新，产生新的父代个体，扩大了种群的多样性，避免算法陷入局部最优。
通过对小规模和大规模测试算例的仿真实验，以及利用改进算法求解具有异质性机器的学习型半导体晶圆制造绿色车间调度问题，验证了本文所提出的算法对于求解具有学习效应的半导体晶圆制造绿色车间调度问题的有效性和可行性。
关键词：学习效应；异质性机器；半导体晶圆制造；绿色调度中图分类号：ＴＰ１８ 文章标识码：Ａ 文章编号：１００７ ３２２１（２０２１）０４ ０２１７ ０７ ｄｏｉ：１０．１２００５／ｏｒｍｓ．２０２１．０１３４ＲｅｓｅａｒｃｈｏｎＧｒｅｅｎＪｏｂＳｈｏｐＳｃｈｅｄｕｌｉｎｇＰｒｏｂｌｅｍｏｆＳｅｍｉｃｏｎｄｕｃｔｏｒＷａｆｅｒｓＭａｎｕｆａｃｔｕｒｉｎｇｗｉｔｈＬｅａｒｎｉｎｇＥｆｆｅｃｔＤＯＮＧＪｕｎ１，２，ＹＥＣｈｕｎ ｍｉｎｇ１（１．ＢｕｓｉｎｅｓｓＳｃｈｏｏｌ，ＵｎｉｖｅｒｓｉｔｙｏｆＳｈａｎｇｈａｉｆｏｒＳｃｉｅｎｃｅ＆Ｔｅｃｈｎｏｌｏｇｙ，Ｓｈａｎｇｈａｉ２０００９３，Ｃｈｉｎａ；２．ＨｅｎａｎＩｎｓｔｉｔｕｔｅｏｆＴｅｃｈｎｏｌｏｇｙ，Ｘｉｎｘｉａｎｇ４５３０００，Ｃｈｉｎａ）Ａｂｓｔｒａｃｔ：Ｉｎｏｒｄｅｒｔｏｍｉｎｉｍｉｚｅｍａｋｅｓｐａｎ，ｔｏｔａｌｃａｒｂｏｎｅｍｉｓｓｉｏｎｓａｎｄｔｏｔａｌｔａｒｄｉｎｅｓｓ，ｗｅｃｏｎｓｔｒｕｃｔａｎｅｗｌｅａｒｎｉｎｇｅｆｆｅｃｔｍｏｄｅｌｗｉｔｈｄｏｕｂｌｅｉｎｆｌｕｅｎｃｉｎｇｆａｃｔｏｒｓｆｉｒｓｔｌｙ．ＴｈｕｓａｎｉｍｐｒｏｖｅｄＭｕｌｔｉ ＶｅｒｓｅＯｐｔｉｍｉｚｅｒａｌｇｏｒｉｔｈｍｉｓｐｒｏｐｏｓｅｄａｎｄｉｔｓｃｏｎｖｅｒｇｅｎｃｅｉｓａｌｓｏｐｒｏｖｅｄ．Ｂｙｐｅｒｆｏｒｍｉｎｇｒｅｖｅｒｓｅｌｅａｒｎｉｎｇｏｎｔｈｅｉｎｉｔｉａｌｐｏｐｕｌａｔｉｏｎ，ｐｅｒｔｕｒｂｉｎｇｔｈｅｉｎｄｉｖｉｄｕａｌｓｉｎｔｈｅｕｎｉｖｅｒｓｅｐｏｐｕｌａｔｉｏｎｂｙＬｅｖｙｆｌｉｇｈｔ，ａｎｄｍｕｔａｔｉｎｇａｎｄｕｐｄａｔｉｎｇｔｈｅｉｎｄｉｖｉｄｕａｌｓｉｎｔｈｅｅｘｔｅｒｎａｌａｒｃｈｉｖｅｂｙｎｅｉｇｈｂｏｒｈｏｏｄｓｅａｒｃｈ，ｎｅｗｐａｒｅｎｔｉｎｄｉｖｉｄｕａｌｓａｒｅｇｅｎｅｒａｔｅｄ，ｗｈｉｃｈｅｘｐａｎｄｔｈｅｄｉｖｅｒｓｉｔｙｏｆｔｈｅｐｏｐｕｌａｔｉｏｎａｎｄａｖｏｉｄｔｈｅａｌｇｏｒｉｔｈｍｆａｌｌｉｎｇｉｎｔｏｌｏｃａｌｏｐｔｉｍｕｍ．Ｔｈｒｏｕｇｈｔｈｅｓｉｍｕｌａｔｉｏｎｅｘｐｅｒｉｍｅｎｔｓｏｆｓｍａｌｌ ｓｃａｌｅａｎｄｌａｒｇｅ ｓｃａｌｅｔｅｓｔｃａｓｅｓ，ａｎｄｔｈｅｕｓｅｏｆｉｍｐｒｏｖｅｄａｌｇｏｒｉｔｈｍｔｏｓｏｌｖｅｔｈｅｇｒｅｅｎｓｈｏｐｓｃｈｅｄｕｌｉｎｇｐｒｏｂｌｅｍｏｆｌｅａｒｎｉｎｇｓｅｍｉｃｏｎ ｄｕｃｔｏｒｗａｆｅｒｓｆａｂｒｉｃａｔｉｏｎｗｉｔｈｈｅｔｅｒｏｇｅｎｅｏｕｓｍａｃｈｉｎｅｓ，ｉｔｉｓｖｅｒｉｆｉｅｄｔｈａｔｔｈｅｐｒｏｐｏｓｅｄａｌｇｏｒｉｔｈｍｆｏｒｓｏｌｖｉｎｇｔｈｅｇｒｅｅｎｓｈｏｐｓｃｈｅｄｕｌｉｎｇｐｒｏｂｌｅｍｏｆｓｅｍｉｃｏｎｄｕｃｔｏｒｗａｆｅｒｓｆａｂｒｉｃａｔｉｏｎｗｉｔｈｌｅａｒｎｉｎｇｅｆｆｅｃｔｉｓｖａｌｉｄａｎｄｆｅａｓｉｂｌｅ．Ｋｅｙｗｏｒｄｓ：ｌｅａｒｎｉｎｇｅｆｆｅｃｔ；ｈｅｔｅｒｏｇｅｎｅｏｕｓｍａｃｈｉｎｅｓ；ｓｅｍｉｃｏｎｄｕｃｔｏｒｗａｆｅｒｓｍａｎｕｆａｃｔｕｒｉｎｇ；ｇｒｅｅｎｓｃｈｅｄｕｌｉｎｇ０ 引言半导体制造系统被誉为当今最复杂的制造系统［１］，在其整个制造过程中以晶圆片的加工过程最为复杂，资金最为密集。
基于文献计量学分析症状群生物学机制的研究现状
ʌ临床基础ɔ基于文献计量学分析症状群生物学机制的研究现状❋李凌香1,刘保延2,周雪忠3,舒梓心3,牟梓君1,何丽云1ә(1.中国中医科学院中医临床基础医学研究所,北京㊀100700;2.中国中医科学院数据中心,北京㊀100700;3.北京交通大学计算机与信息技术学院,北京㊀100044)㊀㊀摘要:目的:分析国内外症状群生物学机制的研究现状及发展态势,为中医药领域开展相关研究提供参考㊂方法:检索7个常用数据库中有关症状群生物机制的文献,采用文献计量学方法分析纳入文献的发表年代㊁研究内容㊁作者㊁高频关键词等㊂结果:共纳入文献175篇,发文量自2012年呈快速增长趋势,美国以34.29%发文量居首位,语种以英文为主(79.43%),以观察性临床研究为主,研究内容主要聚焦于精神心理领域㊁癌症㊁胃肠道疾病㊁慢性疲劳综合征等,研究热点主要集中于抑郁症㊁癌症相关症状群㊁细胞因子及遗传学等方面㊂结论:症状群生物学机制研究已进入大发展时期,但中医相关研究仍处于萌芽阶段,或可借鉴现代医学相关研究成果和经验,结合大数据时代带来的良机,以促进中医药领域相关研究的发展㊂㊀㊀关键词:症状群;生物学机制;生物分子网络;辨证论治;文献计量学㊀㊀中图分类号:R229㊀㊀文献标识码:A㊀㊀文章编号:1006-3250(2021)05-0786-06Research Status of Biological Mechanism of Symptom ClustersBased on Bibliometric AnalysisLI Ling-xiang 1,LIU Bao-yan 2,ZHOU Xue-zhong 3,SHU Zi-xin 3,MOU Zi-jun 1,HE Li-yun 1ә(1.Institute of Basic Research in Clinical Medicine,China Academy of Chinese Medical Sciences,Beijing 100700,China;2.Data center,China Academy of Chinese Medical Sciences,Beijing 100700,China;3.School of Computer and Information Technology,Beijing Jiaotong University,Beijing 100044,China)㊀㊀Abstract :Objective :To analyze the research status and development trend of the biological mechanism of symptom clusters at home and abroad ,so as to provide references for related research in the field of Traditional Chinese Medicine (TCM ).Methods :The data sources were retrieved from 7databases.The key aspects according to bibliometrics method such as publication time ,research content ,author ,high-frequency and keywords were analyzed.Results :A total of 175papers were included ,and the annual volume of papers published has shown a rapid growth trend since 2012;American (34.29%)had the largest amount of documents ,and English was the main language (79.43%);Observational clinical research was the most common type and the research mainly focused on the field of mental psychology ,cancer ,gastrointestinal diseases ,fatigue syndrome ,etc ;It was the hotspot to study depression ,cancer-related symptoms ,cytokines and its genetics.Conclusion :The studies on the biological mechanism of symptom cluster has developed rapidly ,but there was still less on TCM area especially for high quality studies.We can draw lessons from the relevant research and make full use of biomedical database ,and take advantage of the opportunity brought by the era of big data ,which could promote the development of related research on TCM ,and may reveal the scientific connotation of syndrome differentiation and treatment.㊀㊀Key words :Symptom cluster ;Biological mechanism ;Biomolecular network ;Syndrome differentiation treatment ;Bibliometric❋基金项目:国家自然科学基金面上项目(81673964)-基于多尺度非完整数据的中医个体知识图谱构建及其诊疗规律研究作者简介:李凌香(1990-),女,湖南衡阳人,在读博士研究生,从事中医临床评价方法研究㊂ә通讯作者:何丽云(1963-),女,内蒙古锡林郭勒盟人,研究员,博士研究生,硕士研究生导师,从事中医临床评价方法研究,Tel :************,E-mail :Hely3699@ ㊂㊀㊀症状群(symptom cluster )最早出现在心理学和精神卫生领域㊂近10年来,症状群相关研究已成为国内外护理和癌症症状管理的热点,并逐步延伸至慢性疾病[1-4]领域㊂研究表明,单纯通过生物标志物或单个症状对疾病做出诊疗和预后判断并不准确,因此学者们越来越注重对症状群生物学机制的探索㊂2001年,Dodd 等[5-6]首次提出症状群概念:症状群为一组(2个及以上)具有内在联系的症状集合 ㊂中医认为,症状群是指在同一病证中出现,具有共同病理机制且常相伴而生的几个症状共同组成的症状集合[7]㊂国内外学者对 症状群 的认识相对一致,症状群在中医辨证论治中具有重要地位,其本质是医患互动的信息转化过程,而症状群是信息的内涵㊂症状群可以体现疾病不同阶段的特征,用相对固定的 症状群 将 病 的过程㊁变化规律(即病机)加以区分;‘伤寒杂病论“是善于运用症状群的典范,中医先辈们构建了一套阐释方证(症状群)规律的理论与临床诊疗体系,但对症状群生物学机制的阐述和揭示仍存在较大瓶颈㊂本研究分析症状群生物机制的研究现状㊁发展趋势㊁热点及不足,以期为中医药领域开展相关研究提供参考㊂1㊀资料与方法1.1㊀资料来源以中国知识基础设施(china national knowledge infrastructure ,CNKI )㊁中国学术期刊数据库(万方数据)㊁中文科技期刊数据库(维普网)㊁PubMed ㊁SinoMed ㊁Web of science ㊁Ovid ,共7个数据库为资料来源㊂检索主题:症状群生物学机制㊂中文文献检索采用主题/全字段检索,外文文献检索采取主题词与自由词相结合的方式检索,以期尽量查全㊂中文检索式: 症状群 AND 生物标记物 OR 炎症因子 OR 基因 OR 表观基因 OR 代谢组学 OR 蛋白组学 OR 线粒体 OR 细胞因子 OR 细菌 OR 病毒 ㊂外文检索式: Biological Markers [Mesh ]OR biomarker ˑ[Text Word ]OR Inflammation [Mesh ]OR inflammatory markers [Text Word ]OR Epigenesis ,Genetic [Mesh ]OR Genomics [Mesh ]OR Metabolomics [Mesh ]OR Proteomics [Mesh ]OR Mitochondria [Mesh ]OR Cytokines [Mesh ]OR Bacteria [Mesh ]OR Viruses [Mesh ]AND Syndrome [Mesh ]OR symptom cluster [Text Word ]OR symptom clusters [Text Word ]OR symptom complex [Text Word ]OR multiple symptoms [Text Word ][8]㊂Filters :Humans ㊂检索时间范围为建库至2020年3月20日㊂1.2㊀文献纳入标准研究类型涉及临床研究㊁实验研究㊁综述等;研究对象为症状群生物学基础(包括生物标记物㊁炎症因子㊁基因㊁表观基因㊁代谢组学㊁蛋白组学㊁线粒体㊁细胞因子㊁细菌㊁病毒等)㊂1.3㊀文献排除标准动物实验研究㊁无法获取全文的研究㊁数据不完整㊁重复发表或数据重复报道的研究(保留时间最早的1篇)㊁新闻报道㊁社论与主题不相关文献㊂1.4㊀研究质量控制文献筛选过程由2名研究人员根据纳入㊁排除标准,严格独立进行,意见不一致时由双方讨论达成一致意见,或由第三方人员进行判断㊂1.5㊀统计学方法以文献计量学基本定律作为数据分析的基本方法,采用Note-Express 软件与Excel 对相关文献的年代分布㊁语种及国家㊁文献类型㊁作者㊁高频关键词㊁研究内容㊁研究方法等为对象进行统计分析㊂2㊀结果2.1㊀文献检索结果检索到相关文献共8581篇,其中CNKI 40篇,万方108篇,维普7篇,Pubmed 2464篇,Sinomed 2102篇,Web of science 846篇,Ovid 数据库中3014篇㊂将初检文献进行查重,剔除重复文献217篇;阅读文献题目和摘要(必要时阅读全文)后,按照排除标准剔除文献8189篇(排除主题不相关文献较多,原因是现代医学疾病多以 综合征 命名的特点所致,而文献本身并未涉及症状群生物学机制研究),共纳入文献175篇(中文文献30篇,外文文献145篇)㊂2.2㊀文献特征描述2.2.1㊀研究发展趋势分析㊀图1示,检索到有关症状群生物学机制最早的文献见于1983年,发文1篇,表明在 症状群 概念正式提出之前已有学者关注这一问题㊂但1983年至2003年10年间共发表文献21篇,占发文总量的12%,平均每年发表2.1篇,数量少且增长缓慢;从2003年开始小幅度增加,由21篇增加至2012年的71篇,增长了3.4倍㊂自2012年开始呈现快速增长趋势,2012年1月1日至2020年3月20日共检索到发表文献104篇,占发文总量的59.42%,平均每年发表13篇;其中发表文献最多的年份是2019年高达22篇,占总文献数的12.57%,是1983年该类文章首次发文量的22倍㊂普赖斯原理[9]表明,文献数量进入指数增长阶段时,学科便进入大发展时期,提示症状群生物学机制研究的进展较缓慢,但目前已经进入大发展时期,呈现 情报爆炸 态势㊂图1㊀各年度发表文献数量与文献累计数量比较㊀㊀2.2.2㊀文献类型及语种分析㊀文献类型分为期刊㊁学位论文㊁会议论文三类,期刊文章最多为163篇(93.14%),其次为学位论文11篇(6.29%),会议论文1篇(0.57%),表明期刊论文是症状群生物学机制的主要发表形式㊂文献语种为英文论文139篇(79.43%),中文论文30篇(17.14%),其他语种为6篇(3.43%),提示研究症状群生物学机制的论文目前以英文为主㊂2.2.3㊀研究设计类型分析㊀表1示,文献的研究设计类型分为4大类,包括临床研究(病例报告/病例系列㊁横断面/队列研究㊁随机对照试验㊁病例对照)㊁文献综述(传统综述与系统综述)㊁基础实验研究以及其他类型研究㊂以综述类文献68篇(38.86%)与病例对照类论文55篇(31.43%)为多见,表明症状群生物学机制研究目前以观察性临床研究为主,干预性试验研究和基础实验类研究较少㊂表1㊀文献研究设计类型分布比较研究类型论文数量/篇百分比/%基础实验10.57文献综述6838.86病例报告/病例系列2313.14横断面/队列研究2011.43随机对照试验5 2.86病例对照5531.43其他31.71㊀㊀2.2.4㊀文献作者分析㊀表2示,依据洛特卡定律,在一个成熟的研究领域,写n 篇论文的作者数量约为写1篇论文作者数的1/n 2,并且写1篇论文的作者数量约占所有作者数量的60%[10]㊂本研究仅统计论文的第一作者,作者发文量最多为4篇共有5位作者,发文量ȡ2篇的作者共40位㊂表2示,发文量4篇㊁3篇和2篇的作者人数占发文1篇,作者人数的比率分别约为3.70%㊁20.00%和5.93%,比率低于洛特卡定律规定的理论值(分别为1/42㊁1/32和1/22);发表1篇论文作者数约占作者总数的77.14%,高于洛特卡定律规定的60%理论值㊂提示症状群生物学机制相关研究还没有形成成熟的作者群,需要进一步培育能够长期从事该研究领域的学者群㊂表2㊀作者发文量分布比较发文量/篇作者人数占作者总数比率/%占发文1篇作者人数比率/%45 2.86 3.70384.575.9322715.4320.00113577.14100.00㊀㊀图2示,收录文献中以美国作者发文最多共60篇占34.29%,其次为中国39篇占22.29%,英国10篇占5.71%㊂图2㊀作者地域分布情况比较㊀㊀2.2.5㊀研究热点分析㊀表3示,高频关键词是反映研究热点的关键信息,175篇文献共有536个关键词㊂从频次排名前10位的关键词来看,症状群生物学机制的研究热点主要集中于抑郁症㊁细胞因子与癌症相关症状群㊁炎症标记物及遗传学等方面㊂表3㊀排名前10的高频关键词分布序号关键词频次/次频率/%1抑郁症31 5.782细胞因子30 5.603症状群28 5.224炎症23 4.295基因21 3.926癌症14 2.617疲劳12 2.248慢性疲劳综合征11 2.059疼痛10 1.8710人类101.87㊀㊀2.3㊀研究内容分析文献的研究内容能够在一定程度上反映出某时期内某一学科的研究热点及发展方向[11]㊂按照疾病的种类对文献研究内容进行归纳和总结,发现研究内容主要聚焦于7大类疾病,如精神心理疾病(抑郁症㊁精神分裂症㊁围绝经期综合征㊁经前期综合征)㊁癌症相关症状群(放化疗术后及陪护人员出现的症状群)㊁胃肠道症状群(肠易激㊁功能性消化不良)㊁纤维肌痛综合征㊁慢性疲劳综合征㊁心脏疾病症状群(心力衰竭㊁体位性心动过速综合征)㊁干燥综合征㊂分析发现,最早关注症状群生物学机制的文献为外文文献,研究内容为经前期综合征[12]㊂30篇中文文献的主要研究内容为精神心理疾病19篇(63.33%),癌症相关症状群5篇(16.67%),心脏相关疾病症状群4篇(13.33%)㊂145篇英文文献主要研究内容为精神心理疾病36篇(24.83%),癌症相关症状群25篇(17.24%),胃肠道症状群21篇(14.48%),慢性疲劳综合征11篇(7.59%),纤维肌痛综合征11篇(7.59%),干燥综合征8篇(5.52%)㊂2.4㊀研究方法分析2.4.1㊀症状群识别及其分类方法㊀文献中对于命名为 综合征 的疾病,其症状群的识别主要依据临床表现来确定[13];对于癌症相关症状群的识别,多采用潜在剖面分析法(potential profile analysis,LPA)与聚类分析法[14-15],结合患者的基本特征如年龄㊁性别㊁病程㊁病史和使用处方等,采用Logistic回归分析方法对症状群发生发展和预后等因素进行分析,获得影响症状群状态变化的主要表征及干预特征[16]㊂对于精神心理领域症状群的识别,主要根据相关量表(如汉密尔顿抑郁量表(hamilton depression scale,HAMD)㊁安德森症状评估表(md anderson symptom inventory,MDASI)㊁Greene更年期评分量表等)的条目㊁得分以及家族史,进行症状群划分[17-20],提示目前可能尚未形成较统一的症状群识别方法㊂2.4.2㊀生物标志物-基因提取方法㊀共23篇文献涉及基因提取,文献中主要采用聚合酶链反应-限制性片段长度多态性(polymerase chain reaction-restriction fragment length polymorphism,PCR-RFLP)方法(9篇),也有研究者采取高温连接酶检测反应基因芯片(ligase detection reaction,LDR)技术测定基因多态性㊁利用TaqMan荧光探针检测单核苷酸多态性(single nucleotide polymorphism,SNP)技术以及聚合酶链式反应结合变性聚丙烯酰胺凝胶电泳(polymerase chain reaction-polyacrylamide gel electrophoresis,PCR-PAGE)等分析方法,还有少量的(genome-wide association studies,GWAS)研究[21],表明目前临床研究中基因数据主要来源于小样本目标人群的生物样本提取检测,对于生物医学数据库的利用度不高㊂3 讨论3.1㊀症状群生物学机制研究的重要性分析症状群生物学机制研究是中西医结合新医学体系的关键拐点,也是沟通基础研究与临床实践的桥梁㊂症状群生物学机制研究,不仅能为症状群的靶向治疗和精准护理奠定理论基础[22-23],还能揭示中医辨证论治的科学内涵㊂国外1项130万余人的失眠相关基因研究[24]发现,失眠疾病基因与高血压㊁肥胖和冠心病之间有很高的重合度,从某种程度上说明中医异病同治的科学内涵㊂文献研究[25]发现,桂枝汤证的生物学基础为9个toll样受体(toll-like receptors,TLR)的亚型中TLR-3㊁TLR-4和TLR-9的mRNA表达上调,而桂枝汤能有效逆转某些TLR亚型在不同时间的上调,这一研究科学地解释了桂枝汤的药理作用机制,也阐明了桂枝汤证(即症状群)的生物学基础,在一定程度上阐释了 方证 的科学内涵㊂中医的病通常是建立在症状学基础上的,是一种宏观上的病[26]㊂尽管 症 是客观的,但客观性并不是选择诊断性或评价性指标的唯一依据[27],只有能真正反映疾病特征的症状群,才能作为诊断或评价的指标㊂病是症状群在时间上的分布规律的体现,不同时期的症状群将疾病分成段,而证是症状群所处的空间位置㊂症状群是证候形成的主要信息载体,不同类别的症状群可以反映疾病不同阶段的特征㊂用相对固定的 症状群 将 病 的过程㊁变化规律,也就是 病机 加以区分,形成辨证论治诊疗模式的主体框架,可以提高辨证和论治的把握度㊂刘钊乐[28]等发现,症状群具有更强的诊断指向性,对提高临床辨证诊断水平有积极的影响㊂生物学机制研究则是该研究策略下的重要切入点,有助于从分子水平综合描述病证发展过程的复杂性及其相互关联关系[29]㊂在此研究潮流之下,现代医学与中医有望殊途同归,未来在中医辨证论治整体观的指导下,可以系统开展症状表型研究㊁症状群分类㊁生物学基础内涵的研究,建立以症状群为主要框架的个体化辨证论治诊疗体系㊂3.2㊀症状群生物学机制的研究现状分析通过上述文献分析可以发现,症状群生物学机制研究仍处于不成熟阶段㊂现代医学对症状群的研究主要涉及症状群的种类㊁构成及其与基因多态性㊁细胞因子等生物标记物的相关性㊂研究方式以临床观察性研究为主,且以小样本研究居多,干预性试验研究和基础实验类研究仍很少;研究的病种亦较为局限,主要关注精神心理疾病㊁癌症相关症状群㊁胃肠道症状群㊁纤维肌痛综合征与慢性疲劳综合征,未来应该进一步丰富疾病病种,并进一步探索不同病种出现相同症状群的生物学机制㊂目前症状群的识别方法尚未统一,这可能是疾病的命名特点所致㊂现代医学对于许多疾病的命名,尤其是命名为 综合征 的疾病,其本质就是1组或多组由特异性症状与非特异性症状组成的症状群,如肠易激综合征㊁围绝经期综合征㊁干燥综合征等㊂对于症状群的识别,主要使用统计分析方法提取症状群或采用质性研究的描述方法,通过文字提炼症状群[30]㊂但因对症状群定义的界定不一致,研究者采取的统计学方法也存在差异[31]㊂目前常采用的统计学方法主要包括主成分分析㊁公因子分析和聚类分析㊂然而Kim[32]指出,并没有太理想的研究症状群的统计分析方法;2016年Samantha[33]发现,潜类分析法(latent class analysis,LCA)非常适合症状群研究,可用于鉴别不同人群中生物标志物㊁遗传变量与症状群成员的关系㊂因此,研究者在选择识别症状群方法时,应该结合自身研究疾病的特点,选择相对较为公认的方法㊂文献研究提示,目前临床研究中基因数据主要来源于小样本目标人群的生物样本提取检测,对于生物医学数据库的利用度不高㊂因此,笔者建议在开展大型临床研究的同时,可以加大对生物医学数据库的利用度㊂大数据时代,许多开源数据库都可以十分便捷地获取信息和知识㊂本研究团队前期建立的数据库SymMap[34],整合了中西医症状㊁疾病㊁基因㊁中药和靶点等症状图谱(SymMap http:// /)㊂该图谱包含6种实体(1717个中医症状,499味中药,961个西医症状,5235个西医疾病,19595种化学成分和4302个基因)和106721条直接实体关系,且在持续更新和完善中㊂刘林[35]等通过对症状数据的收集整理及研究,得到准确的症状术语㊁症状与疾病以及症状与基因关系数据等,从而为研究症状的生物学基础提供了巨大帮助㊂以下简单介绍一下常用的中医药生物医学数据库㊂①疾病数据库:MalaCards[36]㊁DisGeNet[37]㊁GWAS[38]和DiseaseConnect[39]等;②中药成分数据库:TCMSP[40]㊁TCMID[41]㊁HIT[42]和TCM Database@ Taiwan[43];③药物成分靶点数据库:Genecards[44]㊁DrugBank[45]等;④生物分子相互作用数据库: STRING11(人种子集,近2万蛋白质,30余万高质量相互作用)㊁HPRD㊁BIND㊁DIP㊁HAPPI和MINT[46]等㊂各类生物医学信息数据库正以蓬勃态势发展,为相关研究的开展奠定了深厚的生物学基础㊂3.3㊀症状群生物学机制的研究动态分析2017年美国专家共识[47]提出,在未来的研究中,症状群的定义应该从患者的主观症状体验㊁时间维度㊁基本分子机制等角度出发㊂为进一步揭示症状群的分子机制,促进症状群临床评测的开展,美国NIH于2018年发布了有关症状群研究的R01项目指南(https:///grants/guide/pa-files/ PA-17-462.html)㊂该指南分别从跨病种症状群种类及演变规律㊁症状群的分子机制和症状群的评测研究等三个方面开展,期望通过症状群的临床与基础研究,以解决疾病机制研究中症状涉及少与临床中症状困扰负担重的鸿沟(临床中症状困扰仍是大多数患者就医的主要原因)㊂中医药领域开展症状群生物学机制研究,需要结合中医临床数据在非特异性症状的采集㊁症状内涵的理论解析和辨证论治体系的优势;从疾病的症状群分类㊁症状群生物学内涵和分子网络关联的有效核心方证的挖掘等方面进行示范性研究,为中医药领域防治重大慢性疾病(如糖尿病㊁失眠等)提供可行的方法和研究路径㊂未来开展症状群相关研究,需要注意以下几个方面:一是同时关注特异性症状群和非特异性症状群,以便开展个体化治疗;二是症状群分子机理研究,除特异性生物标记物,还需从分子网络视角探索症状群生物学内涵的系统性与多样性;三是疾病症状群研究,不能只注重单个症状群的发现研究,还需探索各症状群之间的关联及演变规律㊂症状群生物学机制研究目前处于发展阶段,中医疾病-症状群生物学机制研究更是处于萌芽阶段,但大数据时代带来了前所未有的良机,如果能够建立相应的方法结合网络药理学㊁系统生物医学㊁生物信息学㊁互联网技术以及临床诊疗数据,对重大慢性疾病的症状群分类㊁演变规律及生物分子机制进行深入系统的研究,或许有望进一步促进中医辨证论治诊疗规律科学内涵的揭示,这也是中西医结合的一个新的着力点㊂本文通过对症状群生物学机制的研究现状㊁热点与未来可能的发展方向进行分析,结合中医学的学科特点,提出以 症状群 结合 病机 进行疾病分类的设想:通过对疾病-症状群-有效方证为核心的辨证论治体系,展开生物学机制的探索,建立从 疾病-症状群表型网络-中药靶点网络-分子相互作用网络 理解 病-症-治-效 的研究框架,以期为中医药领域开展相关研究提供参考㊂关于构建中医疾病-症状群-有效核心方-分子生物学机制相关联的辨证论治体系,任重而道远,希望有志者能携手共同探索㊂参考文献:[1]㊀MIASKOWSKI C,BARSEVICK A,BERGER A,et al.Advancing symptom science through symptom cluster research:expert panel proceedings and recommendations[J].JNCI:Journal of the National Cancer Institute,2017,109(4):253.[2]㊀LEE S J,JEON J.Relationship between symptom clusters andquality of life in patients at stages2to4chronic kidney disease inKorea[J].Applied Nursing Research,2015,28(4):13-19.[3]㊀CONLEY S,PROCTOR D D,JEON S,et al.Symptom clustersin adults with inflammatory bowel disease[J].Research inNursing&Health,2017,40(5):424-434.[4]㊀MOENS K,SIEGERT R J,TAYLOR S,et al.Symptom Clustersin People Living with HIV Attending Five Palliative CareFacilities in Two Sub-Saharan African Countries:A HierarchicalCluster Analysis[J].Plos One,2015,10(5):e0126554.[5]㊀DODD MJ,MIASKOWSKI C,PAUL SM.Symptom clusters andtheir effect on the functional status of patients with cancer[J].Oncol Nurs Forum,2001,28(3):465-470.[6]㊀DODD M,JANSON S,FACIONE N,et al.Advancing thescience of symptom management[J].Journal of advancednursing,2001,33(5):668-676.[7]㊀郑立群,苏静,杨振宁.症状群落理论对中医辨证思维影响的探究[J].山东中医药大学学报,2016,40(3):208-214.[8]㊀LYNCH KELLY D,DICKINSON K,HSIAO C P,et al.Biological Basis for the Clustering of Symptoms[J].Seminars inOncology Nursing,2016,32(4):351-360.[9]㊀SOLLA PRICE.Little science,big science[J].Von DerStudierstube Zur,1963,7(6):443-458.[10]㊀EGGHE L,ROUSSEAU R.Introduction to Informet rics:Quantitative Methods in Library,Documentation and InformationScience[M].New York:Elsevier Science Publishers,1990:341-344.[11]㊀吴冰.国内糖尿病饮食相关文献计量学分析[J].中国护理管理,2014,14(11):1204-1207.[12]㊀ABRAHAM G E.Nutritional factors in the etiology of thepremenstrual syndromes[J].Journal of Reproductive Medicine,1983,28(7):446-464.[13]㊀AL-DABBAGH M,LAPPHRA K,SCHEIFELE D W,et al.Elevated Inflammatory Mediators in Adults with OculorespiratorySyndrome following Influenza Immunization:a Public HealthAgency of Canada/Canadian Institutes of Health ResearchInfluenza Research Network Study[J].Clinical&VaccineImmunology,2013,20(8):1108-1114.[14]㊀王潇,段培蓓,夏超,等.癌症症状群亚组研究的文献计量学分析[J].护理学报,2019,26(5):20-24. [15]㊀ILLI J,MIASKOWSKI C,COOPER B,et al.Associationbetween pro-and anti-inflammatory cytokine genes and asymptom cluster of pain,fatigue,sleep disturbance,anddepression[J].Cytokine,2012,58(3):437-447.[16]㊀庞秋萍,姬艳博,许翠萍.基于聚类分析的癌症病人症状群及其与生活质量㊁炎症因子的关系[J].护理研究,2019,33(24):4205-4210.[17]㊀刘涛.脑源性神经营养因子基因(GT)n二核苷酸重复多态与精神分裂症的关联研究[D].长沙:中南大学,2004.[18]㊀杨斌.抑郁症患者血浆脑肠肽含量变化及其相关性研究[D].兰州:兰州大学,2005.[19]㊀张晨,吴志国,洪武,等.Bcl-2基因启动子多态性与重度抑郁症及其临床表型的关联分析[J].上海交通大学学报(医学版),2014,34(4):422-425.[20]㊀施梅.单相抑郁症与5-羟色胺系统候选基因的关联研究[D].哈尔滨:哈尔滨医科大学,2006.[21]㊀DANKYU YOON,HYO-JEONG BAN,YOUNG JIN KIM,et al.Replication of genome-wide association studies on asthma andallergic diseases in Korean adult population[J].Bmb Reports,2012,45(5):305-310.[22]㊀MIASKOWSKI C,BARSEVICK A,BERGER A,et al.Advancing Symptom Science Through Symptom Cluster Research:Expert Panel Proceedings and Recommendations[J].J NatlCancer Inst,2017,109(4):1-9.[23]㊀LYNCH KELLY D,DICKINSON K,HSIAO C P,et al.Biological Basis for the Clustering of Symptoms[J].Seminars inOncology Nursing,2016,32(4):351-360.[24]㊀JANSEN PHILIP R,WATANABE KYOKO,STRINGER SVEN,et al.Genome-wide analysis of insomnia in1,331,010individuals identifies new risk loci and functional pathways[J].Nature genetics,2019,51(3):394-403.[25]㊀DU XIN-LIANG,SUI FENG,HUO HAI-RU,et al.Reciprocaleffects of Guizhi decoction to the Guizhi decoction syndrome bytoll-like receptor mRNA expression and cytokines secretion[J].Chinese Journal of Integrative Medicine,2013,19(11):826-835.[26]㊀张宇鹏,尹玉芳.中医 辨证 概念诠释[J].中国中医基础医学杂志,2018,24(10):1352-1353.[27]㊀陈世波,刘保延,王永炎,等. 症 及其分类在中医临床评价中的作用分析[J].中医杂志,2009,50(1):5-7.[28]㊀刘钊乐.中医证候表征方式对辨证诊断思维影响的研究[D].济南:山东中医药大学,2015.[29]㊀李梢.中医证候生物分子网络标志的构想与研究[J].中医杂志,2009,50(9):773-776.[30]㊀CANHUA X,DEBORAH WB,BONNIE MJ.Methods forexamining cancer symptom clustersover time[J].Res NursHealth,2014,37(1):65-74.[31]㊀XIAO C.The state of science in the study of cancer symptomclusters[J].Eur J Oncol Nurs,2010,14(5):417-434. [32]㊀KIM H J,ABRAHAM I,MALONE P S.Analytical methods andissues for symptom cluster research in oncology[J].CurrentOpinion in Supportive and Palliative Care,2013,7(1):45-53.[33]㊀CONLEY S.Symptom Cluster Research With Biomarkers andGenetics Using Latent Class Analysis[J].Western Journal ofNursing Research,2016,39(12):1639-1653. [34]㊀YANG,WU FEILONG,YANG KUO,et al.SymMap:anintegrative database of traditional Chinese medicine enhanced bysymptom mapping[J].Nucleic Acids Res,2019,47(D1):D1110-D1117.[35]㊀刘林.中医临床表型与基因型集成与分析研究[D].北京:北京交通大学,2015.[36]㊀ESPE S.MalaCards:The Human Disease Database[J].Journalof the Medical Library Association JMLA,2018,106(1):140-141.[37]㊀PIÑERO JANET,BRAVOÀLEX,QUERALT-ROSINACH NU'RIA,et al.DisGeNET:a comprehensive platform integratinginformation on human disease-associated genes and variants[J].Nucleic Acids Res,2017,45(D1):D833-D839. [38]㊀VAN,OMMEN,G,et al.Popper revisited:GWAS here,lastyear[J].European Journal of Human Genetics:EJHG,2008,16(1):1-2.[39]㊀CHUN-CHI L,YU-TING T,WENYUAN L,et al.DiseaseConnect:a comprehensive web server for mechanism-based disease-disease connections[J].Nucleic Acids Res,2014,42(Web Server issue):137-146.[40]㊀RU J,LI P,WANG J,et al.TCMSP:a database of systemspharmacology for drug discovery from herbal medicines[J].JCheminform,2014,6(1):13.[41]㊀XUE R,FANG Z,ZHANG M,et al.TCMID:traditionalChinese medicine integrative database for herb molecularmechanism analysis[J].Nucleic Acids Res,2013,41(Database issue):D1089-D1095.[42]㊀YE H,YE L,KANG H,et al.HIT:linking herbal activeingredients to targets[J].Nucleic Acids Res,2011,39(Database issue):D1055-D1059.[43]㊀CHEN CY.TCM Database@Taiwan:the world s largesttraditional Chinese medicine database for drug screening in silico[J].PLoS One,2011,6(1):e15939.[44]㊀REBHAN M,CHALIFACASPI V,PRILUSKY J,et al.GeneCards:integrating information about genes,proteins anddiseases[J].Trends in Genetics Tig,1997,13(4):163. [45]㊀WISHART DS,KNOX C,GUO AC,et al.DrugBank:aknowledgebase for drugs,drug actions and drug targets[J].Nucleic Acids Res,2008,36(Database issue):D901-D906.[46]㊀张彦琼,李梢.网络药理学与中医药现代研究的若干进展[J].中国药理学与毒理学杂志,2015,29(6):883-892.[47]㊀MIASKOWSKI C,BARSEVICK A,BERGER A,et al.Advancing symptom science through symptom cluster research:expert panel proceedings and recommendations[J].Journal ofthe National Cancer Institute,2017,109(4):1-9.收稿日期:2020-06-10。
论文参考文献著录格式规范
论文参考文献著录格式规范参考文献著录格式执行GB/T 7714-2019《信息与文献参考文献著录规则》，采用顺序编码制。
常见的参考文献类型著录格式如下。
1 专著以单行本或多卷册(在限定的期限内出齐)形式出版的印刷型或非印刷型出版物，包括普通图书、古籍、学位论文、会议文集、汇编、标准、报告、多卷书、丛书。
1.1 著录格式：主要责任者.题名：其他题名信息[文献类型标识/文献载体标识].其他责任者.版本项.出版地：出版者，出版年：引用页码[引用日期].获取和访问路径.数字对象唯一标识符.1.2 示例[1] 刘国钧,郑如斯.中国书的故事[M].2版.北京：中国青年出版社,1979：115.[2] 陈登原.国史旧闻：第1卷[M].北京：中华书局,2019：29.[3] Seth A. Spector. 外科学[M].刘淑芬,译.天津：天津科技翻译出版公司,2019：16.[4] 赵学功.当代美国外交[M/OL].北京：社会科学文献出版社,2019：5-6[2019-11-01]. 33023884/1.[5] 王清任.医林改错[M].李天德,整理.北京：人民卫生出版社,2019：68.[6] 王夫之.宋论[M].刻本.金陵：湘乡曾国荃,1865(清同治四年).[7] 张筑生.微分半动力系统的不变集[D].北京：北京大学数学研究所, 1983.[8] 吴云芳.面向中文信息处理的现代汉语并列结构研究[D/OL].北京：北京大学,2019[2019-10-14].[9] 辛希孟.信息技术与信息服务国际研讨会论文集：A集[C].北京：中国社会科学出版社,1994.[10] 陈志勇.中国财税文化价值研究：“中国财税文化国际学术研讨会”论文集[C/OL].北京：经济科学出版社,2019[2019-10-14].http://apabi./usp/pku/pub.mvc?pid=book.detail&metaid=m.20190628-BPO-889-0135&cult=CN.[11] 全国文献工作标准化技术委员会第七分委员会.中国标准书号：GB/T 5795-1986[S].北京：中国标准出版社,1996：2.[12] 冯西桥.核反应堆压力管道与压力容器的LBB分析[R].北京：清华大学核能技术设计研究院,1997.[13] 中国第一历史档案馆,辽宁省档案馆.中国明朝档案总汇[A].桂林：广西师范大学出版社,2019.2 专著中的析出文献2.1 著录格式：析出文献主要责任者.析出文献题名[文献类型标识/文献载体标识].析出文献其他责任者//专著主要责任者.专著题名：其他题名信息.版本项.出版地：出版者，出版年：析出文献的页码[引用日期].获取和访问路径.数字对象唯一标识符.2.2 示例[1] 罗云.安全科学理论体系的发展及趋势探讨[M]//白春华,何学秋,吴宗之.21世纪安全科学与技术的发展趋势.北京：科学出版社,2019：1-5.[2] 贾东琴,柯平.面向数字素养的高校图书馆数字服务体系研究[C]//中国图书馆学会.中国图书馆学会年会论文集：2019年卷.北京：国家图书馆出版社,2019：45-52.[3] ROBERSON J A, BURNESON E G. Drinking water standards, regulations and goals[M/OL]//American Water Works Association. Water quality & treatment: a handbook on drinking water. 6th ed. New York: McGraw-Hill, 2019:1-36[2019-12-10]. Open.aspx?id=291430.3 连续出版物中的析出文献连续出版物是指载有年卷期号或年月日顺序号，并按计划无限期连续出版发行的出版物，包括期刊、报纸。
APPLICATION RUNTIME DETERMINED DYNAMICAL ALLOCATIO
专利名称：APPLICATION RUNTIME DETERMINEDDYNAMICAL ALLOCATION OFHETEROGENEOUS COMPUTE RESOURCES 发明人：LIPPERT, Thomas,FROHWITTER, Bernhard 申请号：EP2019/051615申请日：20190123公开号：WO2019/145354A1公开日：20190801专利内容由知识产权出版社提供专利附图：摘要：The present invention provides a method of operating a heterogeneous computing system comprising a plurality of computation nodes and a plurality of boosternodes, at least one of the plurality of computation nodes and plurality of booster nodes being arranged to compute a computation task, the computation task comprising a plurality of sub-tasks, wherein in a first computing iteration, the plurality of sub-tasks are assigned to and processed by ones of the plurality of computation nodes and booster nodes in a first distribution; and information relating to the processing of the plurality of sub-tasks by the plurality of computation nodes and booster nodes is used to generate a further distribution of the sub-tasks between the computation nodes and booster node for processing thereby in a further computing iteration.申请人：PARTEC CLUSTER COMPETENCE CENTER GMBH地址：Possartstr. 20 81679 München DE国籍：DE代理人：TOMLINSON, Edward James更多信息请下载全文后查看。
System and method for narrowband pre-detection sig
专利名称：System and method for narrowband pre-detection signal processing for passivecoherent location applications发明人：Kevin W. Baugh,Robert H. Benner申请号：US10878170申请日：20040628公开号：US07019692B2公开日：20060328专利内容由知识产权出版社提供专利附图：摘要：A system and method for narrowband pre-detection signal processing inpassive coherent location applications is disclosed. A receiving subsystem receives areference signal and a target signal from an uncontrolled transmitter. The target signal is reflected from a target. The passive coherent location system includes subprocessors that perform pre-detection operations on the reference and target signals. The functions include zero-doppler cancellation, quadrature demodulation, reference beam regeneration, coherent processing interval selection, power spectral density estimation, cross ambiguity function formation, and the like. Within these operations, the reference signal is filtered with respect to the target signal to form a first output reference signal. The first output reference signal is combined with the first target signal to form a first output target signal. The output target signal then is used for subsequent passive coherent location processing operations. The filter is updated with respect to a difference between the target signal and a subsequent target signal. Further, two paths are used for correlation processing of the reference and target signals.申请人：Kevin W. Baugh,Robert H. Benner地址：Gaithersburg MD US,Gaithersburg MD US国籍：US,US代理机构：Marsh Fischmann & Breyfogle LLP更多信息请下载全文后查看。
单细胞基因组测序技术新进展及其在生物医学中的应用
Hereditas (Beijing) 2021年2月, 43(2): 108―117 收稿日期: 2020-12-01; 修回日期: 2021-01-03基金项目:中国博士后科学基金面上项目(编号：2019M651377)和上海市“超级博士后”激励计划项目(2018-2020)资助[Supported by ChinaPostdoctoral Science Foundation Grant (No. 2019M651377), and Shanghai “Super Postdoctoral Fellow” Program (2018-2020)]作者简介: 王卓，博士，研究方向：循环肿瘤细胞鉴定与单细胞测序。
E-mail:*********************.cn 通讯作者:施奇惠，博士，研究员，研究方向：液态活检与单细胞分析。
E-mail:******************.cn DOI: 10.16288/j.yczz.20-363 网络出版时间: 2021/1/22 10:34:26URI: https:///kcms/detail/11.1913.R.20210122.0900.002.html综 述单细胞基因组测序技术新进展及其在生物医学中的应用王卓，申笑涵，施奇惠复旦大学生物医学研究院, 上海 201100摘要: 随着单细胞基因组测序技术的建立与发展，对细胞基因组特征的分析进入了单细胞水平。
单细胞的基因组分辨率不但使研究人员能够在单细胞尺度上分析肿瘤细胞的异质性，也使得传统上难以检测的稀有细胞的基因组研究成为可能。
这些稀有细胞往往具有重要的生物学意义或临床价值，如癌症患者血液中循环肿瘤细胞(circulating tumor cell, CTC)的基因组检测或三代试管婴儿植入前胚胎细胞的遗传缺陷诊断与筛查(preim-plantation genetic diagnosis/screening, PGD/PGS)。
基于智能手机传感器的人类行为识别研究
doi:10.3969/j.issn.1003-3114.2023.03.023引用格式:丁阁文,丁绪星,许蓉,等.基于智能手机传感器的人类行为识别研究[J].无线电通信技术,2023,49(3):566-576.[DING Gewen,DING Xuxing,XU Rong,et al.Research on Human Activity Recognition Based on Smartphone Sensors[J].Radio Com-munications Technology,2023,49(3):566-576.]基于智能手机传感器的人类行为识别研究丁阁文,丁绪星∗,许㊀蓉,王㊀冲,邹孝龙(安徽师范大学物理与电子信息学院,安徽芜湖241002)摘㊀要:针对因人类行为识别中手机位置的不确定性导致的行为识别率低的问题,提出了一种实时行为识别方法㊂首先对智能手机中加速度计和陀螺仪所获取的各行为下的传感器数据,依次进行零偏误差补偿㊁异常值处理和db7小波阈值去噪;其次使用滑动窗口对预处理后的数据进行分割,同时基于各行为特征分布规律进行特征提取;最后将提取的特征用于支持向量机(Support Vector Machine,SVM)分类器模型的训练,通过调用SVM 模型结合多票判决策略实现当前时刻行为预测㊂仿真结果表明,该方法能有效地识别坐㊁站立㊁行走㊁跑步㊁上楼㊁下楼㊁躺㊁卧8种基本行为,并能区分握㊁阅读㊁接听㊁摆臂㊁口袋5种手机位置㊂分类器模型的整体识别率达到了97.6%,在手机位置固定和不固定的情况下,实时行为识别的平均识别准确率分别达到了90.6%和84.8%,平均耗时为177.803ms㊂关键词:人类行为识别;智能手机传感器;支持向量机;小波分析中图分类号:TP391.4㊀㊀㊀文献标志码:A㊀㊀㊀开放科学(资源服务)标识码(OSID):文章编号:1003-3114(2023)03-0566-11Research on Human Activity Recognition Based on Smartphone SensorsDING Gewen,DING Xuxing ∗,XU Rong,WANG Chong,ZOU Xiaolong(School of Physics and Electronic Information,Anhui Normal University,Wuhu 241002,China)Abstract :There is low activity recognition rate caused for the uncertainty of smartphone positions in human activity recognition.Aiming at above problem,a real-time activity recognition method is proposed in this paper.First,the raw data is processed using zerobias error compensation,outlier processing and db7wavelet threshold denoising respectively,which is the sensor data of various behav-iors obtained by the accelerometer and gyroscope of smart phone.Then,the processed data is segmented by a sliding window.At the same time,the features are extracted based on data feature distribution law of each activity.Finally,Support Vector Machine(SVM)al-gorithm and the extracted features are used for training of the classifier model.Behavior is predicted by invoking the SVM model com-bined with the multi-vote decision strategy at the current moment.Simulation results show that eight main behaviors (sitting,standing,walking,running,going upstairs,going downstairs,half lying,and lying)under five smartphone positions (holding,reading,answering,arm swinging,pocket)can be effectively identified.The overall recognition rate of the classifier model reached 97.6%.When the mobilephone position is fixed and not fixed,the average recognition accuracy of real-time behavior recognition reaches 90.6%and 84.8%,re-spectively,and the average time-consuming is 177.803ms.Keywords :human activity recognition;smartphone sensors;SVM;wavelet analysis收稿日期:2023-01-15基金项目:安徽省发展和改革委员会支持项目(832132)Foundation Item :Anhui Development and Reform Commission (832132)0 引言近年来,人类行为识别(Human Activity Recog-nition,HAR)成为了人工智能领域的研究热点,并被广泛地应用于老年监护[1-2]㊁智慧医疗[3]㊁运动监测[4]等领域㊂HAR 的研究方法大致可分为两类:基于计算机视觉的行为识别[5]和基于可穿戴设备的行为识别[6-7]㊂基于计算机视觉的行为识别是利用图像或视频来提取人体的行为特征,这种方法通常要求固定照相机㊁摄像机等硬件设备,且在使用中视线容易被障碍物遮挡,存在局限性[8]㊂而基于可穿戴设备的行为识别主要使用设备中的惯性传感器㊁心率传感器等获得人体行为数据和生理信息,研究人员利用这些数据来预测人们站立㊁行走㊁跑㊁上下楼㊁跳跃等基本行为[9-10]㊂因此与基于计算机视觉的行为识别相比,可穿戴设备方便携带㊁成本低㊁灵敏性高,能方便高效地收集数据且数据处理较简单,在行为识别领域应用更为广泛㊂在过去的10年中,智能手机已经成为一些可穿戴传感器的替代品㊂如今的智能手机中搭载了各种类型的传感器,如GPS㊁惯性传感器(加速度计㊁磁力计㊁陀螺仪)和环境传感器(麦克风㊁光线传感器㊁距离传感器)等[11]㊂不仅如此,智能手机还具有强大的计算能力,可以轻松高效地收集和处理有关用户的丰富信息㊂相对于可穿戴设备,使用智能手机进行数据采集与处理更加方便㊁更具便携性,不管是将手机拿在手里还是装在口袋里,智能手机均可以持续地感知用户的行为㊂尤其是在COVID-19 (Coronavirus Disease2019)较为严重的情况下[12],基于智能手机的行为识别成为了一种高效㊁廉价和安全的方法,利用智能手机传感器数据预测人类行为从而实现对身体行为变化和心理健康的监测,而不需要专门的或者昂贵的医疗备来监测㊂因此,基于智能手机传感器的行为识别具有更高的研究价值㊂1㊀当前研究现状近年来,尽管研究者们在利用智能手机传感器识别人类行为的研究中取得了一些成果,但仍存在以下不足:①手机放置的位置对识别结果有着很大的影响[13]㊂研究者们通常要求将手机放在身体上的特定位置(比如腰部[13]㊁口袋[14]等),研究的实验位置是不自然的,携带不方便;②加速度数据对多种行为数据难以区分,尤其是在识别上下楼梯时,结果不太理想;③大部分的研究者只是致力于数据分析及处理,未结合实际做实时行为识别的研究㊂针对以上问题,Menhour等人[15]使用来自文献[16]的公开在线数据集,将加权K近邻(Weigh-ted K-Nearest Neighbors,WKNN)算法和支持向量机(Support Vector Machine,SVM)相结合,对6项人体行为(行走㊁站㊁坐㊁跑㊁上楼㊁下楼)和4种手机位置(右牛仔裤口袋㊁手臂㊁腰带和手腕)进行识别;其结果为:手机固定在腰部时,上㊁下楼的识别率分别在91.8%和94.2%左右;手机固定在手腕时,上㊁下楼的识别率分别达到了98.1%和99.6%㊂虽然识别率有所提高,但在正常生活中,手机放置在这两处的较少㊂Wang等人[17]在SVM算法基础上引入了D-S 证据理论,再利用数据融合的思想识别人类行为;使用智能手机获取4种行为(行走㊁跑㊁上楼和下楼)以及4种手机位置(阅读㊁接听㊁摇摆和口袋)下的加速度计和陀螺仪数据;该研究中将跑的识别准确率提高到了99.7%,但上楼的准确率仅有92.1%㊂孔菁等人[18]同样使用智能手机的加速度计和陀螺仪数据并采用主成分分析方法对数据统计特征进行降维处理,将决策树与SVM相结合实现对行为特征的分类,模型对所有行为识别的平均准确率达到97.5%,上下楼的识别率也能在96%左右;但文章未给出手机放置位置㊂Qi等人[19]提供了一个基于智能手机的自适应人类行为识别实时监控系统识别12种人类行为,其中包括5种动态行为(步行㊁慢跑㊁跳跃㊁上楼和下楼)㊁6种静态行为(站立㊁坐㊁左右躺㊁俯卧和仰卧)以及一系列的过渡行为(如从坐到站)㊂在模型训练和测试过程中,使用了4种携带手机的方式:裤子的口袋㊁衬衫的口袋㊁腰部和背包,并且将手机按任意方向携带,模型对手机置于腰部和口袋的行为识别取得较好的性能(腰部95.15%,口袋92.20%)㊂在最后的实时行为识别实验中,仅验证了手机放在腰部和左边口袋的两项新的行为:深蹲㊁扭屁股㊂针对以上问题,考虑实时的人类行为识别,本文对5种手机位置下(握㊁阅读㊁接听㊁摆臂㊁口袋)的8种人类基本行为(坐㊁站立㊁行走㊁跑㊁上楼㊁下楼㊁躺㊁卧)展开了研究,并设计了一种基于SVM结合多票判决的实时行为识别方法㊂为减少手机位置对识别率的影响,将行为识别分为离线阶段和在线阶段,主要研究内容包括:数据采集㊁数据预处理㊁特征提取与特征选择㊁分类器模型的训练以及实时行为识别设计5个方面㊂在离线阶段,数据经过预处理后提取相关特征作为输入向量用来训练分类器模型;在线阶段中,对逐渐累积的数据采用离线阶段的数据处理方法进行特征提取,然后调用分类器并结合多票判决的方法预测当前行为㊂2㊀方法框架概述按照人类行为识别研究的总体框架,将实时行为识别分为两个阶段:离线阶段和在线阶段㊂离线阶段具体流程如图1所示,包括:数据采集㊁数据预处理㊁特征提取㊁特征选择以及分类器模型的训练㊂首先,以50Hz的采集频率获取各行为下智能手机中加速度计和陀螺仪的数据,并对采集到的数据进行预处理,依次为零偏误差补偿㊁基于箱型图和均值修正法的异常数据处理㊁以及基于db7小波基和SURE 阈值的软阈值去噪㊂其次,采用固定长度为64,重叠率为50%的滑动窗口对与处理后的数据进行划分,在每个窗口中对数据提取特征用于描述每种行为,并对提取的特征进行筛选以减少冗余特征的影响㊂最后将特征选择后形成的新数据集用于分类器模型的训练㊂图1㊀离线阶段主要流程Fig.1㊀Main process in offline stage在线阶段主要实现实时行为识别,流程如图2所示㊂图2㊀在线阶段主要流程Fig.2㊀Main process of online stage此阶段仍采用固定长度为64的窗口对采集的数据进行划分,当从开始采集数据时长达到1.28s时,启动实时行为识别算法㊂随着数据的增加,每0.2s(数据集S 长度Len 每增加10)更新一次窗口,同时对当前窗口的数据采用与离线阶段相同的预处理,并按照离线阶段中最终选取的特征提取特征向量,再调用训练好的SVM 分类器预测此时行为㊂由于超过1.28s 后的数据会被划分到多个窗口中,可能被标记为不同的行为,此时结合多票判决策略[20],选择出当前位置最终的结果㊂3㊀数据采集与处理3.1㊀数据采集与数据分割将待识别的行为划分为两大类:主要行为和次要行为㊂主要行为是指人的身体姿势或运动状态,本文研究围绕人类日常的基础行为,包括静态动作和动态动作㊂静态动作包括:站㊁坐㊁躺㊁卧;动态动作包括:行走㊁跑㊁上楼和下楼㊂而次要行为是指智能手机的位置㊂根据对人类日常行为的分析,在数据采集过程中分别将手机放置在如图3所示的5个位置上㊂智能手机在位置1~5中分别表示手持不摆臂㊁阅读手机㊁接听电话㊁放置口袋中和手持摆臂,而每一种放置位置都对应人类日常的几种主要行为,如在位置3接听电话的前提下,采集对象可以处于行走㊁跑㊁上楼㊁下楼㊁站㊁坐㊁卧㊁躺这几种行为状态㊂即在不同的位置下完成对不同基本行为的数据采集㊂因此本文所研究的行为如表1所示㊂图3㊀数据采集手机位置示例图Fig.3㊀Sample graph of data acquisition mobile phone location表1㊀待研究行为类别Tab.1㊀Categories of behavior to be studied次要行为主要行为位置1行走㊁跑㊁上楼㊁下楼㊁站位置2行走㊁跑㊁上楼㊁下楼㊁站㊁坐㊁卧㊁躺位置3行走㊁跑㊁上楼㊁下楼㊁站㊁坐㊁卧㊁躺位置4行走㊁跑㊁上楼㊁下楼㊁站㊁坐位置5行走㊁跑㊁上楼㊁下楼㊀㊀实验所用的数据均使用华为nova4安卓手机作为数据采集设备,并以50Hz的频率采集数据,即每20ms获取一组传感器数据㊂因此每一次采集的数据包括7列测量值,其中第一列为时间,第2~7列为分别是加速度计和陀螺仪的测量数据㊂在数据处理之前,先使用滑动窗口算法对数据进行划分,在数据分割时每个窗口内尽可能包含一个及一个以上的完整周期,以保证每个窗口内提取的数据特征能够准确地反映某种行为㊂为选择合适的窗口长度,实验人员在较慢速度的情况下采集一些动态行为数据,通过计算得到行走㊁跑㊁上楼和下楼的平均周期分别为1.04㊁0.65㊁1.18㊁1.04s㊂因此,本文通过固定长度为64(1.28s),50%重叠率的滑动窗口来分割数据㊂所以,每个窗口内的数据样本通过64次采集得到,且每一次采集均可获得3个加速度数据和3个角速度数据,故数据分割后得到的每个数据样本由384个数据组成㊂3.2㊀数据预处理3.2.1零偏误差补偿行为识别对于传感器的精度要求相对较低,所以本文对传感器零偏误差仅做粗略估计㊂零偏误差包含了常值零偏㊁零偏稳定性和零偏不稳定性㊂对零偏误差的处理分为两步:①通过传感器在初始启动过程中采集几秒钟的静态数据求得的平均值来补偿常值零偏和零偏稳定性,测算结果如表2所示㊂②通过N秒平均法测算传感器的零偏不稳定性,如表3和表4所示㊂N秒平均法,即采集几个小时的静态数据,每10s或100s求取平均,最后统计这些平均值的均值与标准差㊂表2㊀静态数据求零偏不稳定性测量结果Tab.2㊀Static data to obtain zero bias instabilitymeasurement results坐标轴加速度计/(m/s2)陀螺仪/((ʎ)/s)X0.0946-0.0008Y-0.00200.0029Z0.0492-0.0082表3㊀N秒平均法求零偏不稳定性测量结果Tab.3㊀N-second average to obtain zero bias instability measurement results坐标轴加速度计/(m/s2)陀螺仪/((ʎ)/h)X-0.013611.6017Y0.12149.4827Z-0.0650 6.2797表4㊀传感器零偏不稳定性的N秒平均法测量结果标准差Tab.4㊀Standard deviation of sensor zero bias instability measured by N-second mean method传感器N=1N=10N=100加速度计X轴/(m/s2)0.006610.006330.00598加速度计Y轴/(m/s2)0.021180.021070.02113加速度计Z轴/(m/s2)0.060390.058790.05834陀螺仪X轴/((ʎ)/h) 2.65370 2.62860 2.62970陀螺仪Y轴/((ʎ)/h)16.0747016.0710016.14170陀螺仪Z轴/((ʎ)/h) 2.19210 2.17080 2.17390 3.2.2异常数据处理本文对异常值的处理采用均值修正法,利用原异常值所在位置的左右各两个数据的平均值来替换该异常值㊂但在处理过程中,需要注意判断数据是否越界㊂当进行异常值处理时,先对去除零偏误差后的数据采用箱型图检测出异常值所在位置㊂若异常值所在位置的左右各两个值均在原数据可检索范围内,则正常进行均值修正㊂若在寻找异常值相邻数据时,左边或右边的相邻数据索引超出了原数据序列长度,此时,只选择可检索到的数据求平均值来替换该异常值㊂图4列举了一组行为加速度数据异常值的处理结果,图中的点代表数据中的异常值㊂从图中可以看出,经处理后异常值数量明显减少㊂经大量的实验仿真,对于异常值的处理方法同样适用于其他行为数据㊂图4㊀异常数据处理前后箱型图对比Fig.4㊀Comparison of box type before and afterabnormal data processing3.2.3小波阈值去噪对于原始数据中存在的高频噪声,本文研究采用小波阈值去噪方法加以滤波,该方法不仅能够剔除信号中的噪声,而且可以尽可能地保留原始数据的特征㊂小波阈值去噪方法的主要流程如图5所示,主要步骤如下:①选择一种小波基函数并确定分解层次;②对含有噪声的数据进行小波变换,以获得不同尺度下的小波系数;③选择合适的阈值,将噪声产生的小波系数删除,只保留真实信号所产生的系数;④对处理后的小波系数利用小波逆变换实现信号重构,以获得去噪后的信号㊂在以上过程中,每个部分的设计,均会直接影响对数据的去噪效果㊂图5㊀小波阈值去噪的流程图Fig.5㊀Flowchart of wavelet threshold denoising在小波去噪中,通常使用信噪比和均方根误差来评价信号的去噪效果㊂信噪比是指信号和噪声二者间能量的比值,计算公式如式(1)所示㊂式中, f(x)为原始信号,f d(x)为去噪后的信号㊂均方根误差用于判断去噪前后信号的分散程度,计算公式如式(2)所示㊂SNR=ðN i=1f2(x i)ðN i=1[f(x i)-f d(x i)]2,(1)RMSE=1NðN i=1[f(x i)-f d(x i)]2㊂(2)一般地,信噪比越大,均方根误差越小,说明去除的噪声越多,信号去噪前后的偏差越小,去噪效果越好㊂依据上述两个指标,在实验仿真中反复调整各参数,选择出适应于大部分行为数据去噪的参数㊂图6和图7中分别展示了一组静态行为数据和一组动态行为数据的去噪前后对比图㊂经过大量的仿真分析确定了在小波阈值去噪处理中,均采用db7小波基函数㊁分解层数为3层结合基于SURE的自适应软阈值去噪方法对各行为数据进行去噪,以提高行为识别的精度㊂图6㊀行为 行走 加速度数据去噪前后对比图Fig.6㊀Comparison of acceleration data before and after denoising of walking图7㊀行为 站 加速度数据去噪前后对比图Fig.7㊀Comparison of acceleration databeforeand after denoising of station3.3㊀特征提取与特征选择为了提高预测的准确性,利用特征提取与特征选择来构建更快㊁消耗更低的预测模型㊂特征提取和特征选择通过选择重要特征的子集来减少数据的维度,以增强模型的可解释性㊂在3.1节的基础上,在每个数据样本中,利用采集到的三轴加速度数据计算出合加速度,并将其作为该数据样本的第7列测量值㊂对于每个窗口内的数据可按照表5所列的特征对7列测量值分别提取19个时域特征和16个频域特征,另外还需提取加速度计三轴测量值之间的皮尔逊相关系数(3个),陀螺仪三轴测量值之间的皮尔逊相关系数(3个),样本数及分布如表6所示㊂表5㊀时域特征㊁频域特征提取Tab.5㊀Time domain feature extraction and frequency domain feature extraction数据时域特征频域特征各列测量值㊁合加速度最小值㊁最大值㊁均值㊁众数㊁四分位数(3个)㊁四分位距㊁极差㊁方差㊁标准差㊁均方根㊁绝对值的平均值㊁平均绝对离差㊁波形因子㊁峰值因子㊁裕度因子㊁脉冲因子㊁能量最小值㊁最大值㊁均值㊁四分位数(3个)㊁四分位距㊁极差㊁方差㊁标准差㊁均方根㊁峰度㊁偏度㊁平均绝对离差㊁平均频率㊁能量加速度计三轴之间皮尔逊相关系数(3个值)无陀螺仪三轴之间皮尔逊相关系数(3个值)无表6㊀不同手机位置下行为的样本数Tab.6㊀Number of samples of behavior at differentphone locations主要行为位置1位置2位置3位置4位置5行走13201335134313211315跑12971296128512941259上楼12391279124312311258下楼12421290124312361252站1311130612981313坐129512761309躺12231223卧12261223㊀㊀为了提高各行为的识别准确率㊁缩短模型的训练时间,使分类器模型具有更好的性能,需要在提取的特征中选择出有效的特征,且这些特征易于区分㊁便于提取㊁数量尽可能少㊂本文根据特征分布规律来选择特征㊂图8为本文8种主要行为合加速度相关特征的分布统计图㊂图8(a)为随机抽取的10000个静态行为样本和动态行为样本的合加速度幅值的最大值分布情况,从图中可以看出,静态行为的合加速度最大值大部分集中在9.6~13m/s2,而动态行为则集中分布在10~80m/s2㊂图8(b)和图8(c)分别为4种动态行为和4种静态行为的合加速度幅值的均值分布情况,对于4种静态行为,合加速度幅值大致分布相同;在4种动态行为中, 跑 的数据特征分布明显与另外3种行为的特征分布有所差异㊂因此,可使用合加速度幅值的相关特征来区分静态行为㊁ 跑 和另外3种动态行为㊂(a)合加速度最大值分布(b)各静态行为合加速度均值分布(c)各动态行为合加速度均值分布图8㊀8种主要活动合加速度幅值相关特征分布Fig.8㊀Characteristic distribution of amplitude correlation of the combined acceleration of eight majoractivities㊀㊀为区分除 跑 以外的另外3种动态行为,以位置4下的各行为数据为例㊂图9中列出了位置4时 行走 上楼 下楼 加速度计X 轴和Z 轴的最小值和最大值的详细分布情况㊂从图中可知,虽然3种行为的数据分布出现重叠,但也存在一些差异,图9(a)和图9(b)中以最上面的虚线为参考面, 行走 的加速度计X 轴和Z 轴的最大值75%以上的数据在虚线上方,而 下楼 的数据特征有50%以上在虚线以下;图9(a)和图9(b)中以最下面的虚线为参考面, 上楼 下楼 两种行为的加速度计X 轴和Z 轴的最小值75%以上的数据在虚线上方,而 行走 的数据特征接近75%的数据分布在虚线以下㊂(a )加速度计X轴最值分布(b )加速度计Z 轴最值分布图9㊀位置4下 行走 上楼 下楼 加速度计最值分布Fig.9㊀Maximum accelerometer value distribution ofwalk , go up and go down at position 4对于同一种主要行为,不同手机位置下的数据特征也有所差异,以 行走 为例,图10分别列出了各种手机位置下加速度计Y 轴㊁Z 轴的数据特征分布㊂从图10(a)可以看出,位置1与位置2㊁位置3相比,数据特征分布有着较明显的差异;图10(b)中,位置2的数据与其他位置上的数据也有着明显的划分界限㊂(a )加速度计Y轴最值分布(b )加速度计Z 轴最值分布图10㊀不同手机位置下 行走 的加速度计三轴最值分布Fig.10㊀Accelerometer triaxial maximum value distributionof walk under different mobile phone positions按照上述方法,对提取的特征进行筛选㊂在原特征集的基础上选择了合加速度部分特征,加速度计和陀螺仪各轴最值㊂由于在不同的行为中,加速度计的三轴数据之间,以及陀螺仪的三轴数据之间有一定的相关性㊂因此,特征选择时保留加速度计与陀螺仪三轴之间的皮尔逊相关系数㊂本文最终选取了30个特征组成特征子集,分别为:加速度计和陀螺仪各轴测量值的最大值㊁最小值以及三轴数据之间的皮尔逊相关系数,合加速度的最小值㊁最大值㊁方差㊁平均数㊁众数㊁四分位数㊁极差㊁四分位距㊁平均绝对离差以及信号能量㊂4㊀模型训练4.1㊀SVM 分类器模型SVM 算法在非线性问题上有着突出的表现,它通过非线性变换将特征向量转换到高维的特征空间中,在高维特征空间中建立一个最大间隔的超平面来实现训练样本分类,巧妙地将非线性问题转化为线性问题㊂最终的超平面对应的模型为:f(x)=w Tφ(x)+b,(3)式中,法向量w和位移项b需满足式(4)所示的目标函数㊂由于求解映射到高维空间后的关系式较为困难,此时可以利用拉格朗日对偶性转换成求解其对偶变量的优化,对偶问题的目标函数如式(5)所示:min w,b12 w 2,㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀s.t.㊀y i(w Tφ(x)+b)ȡ1,㊀i=1,2, ,m,(4) maxαðm i=1αi-12ðm i=1ðm j=1αiαj y i y jφ(x i)Tφ(x j), s.t.㊀ðm i=1αi y i=0,㊀αiȡ0,i=1,2, ,m㊂(5)式(5)中的计算涉及到φ(x i)Tφ(x j)的计算,即样本x i和x j映射到新的特征空间后的内积㊂由于经过变换后的特征空间维数可能很高,所以很难直接对公式展开计算㊂假设一个函数K(x i,x j)满足式(6),则只需通过函数K(x i,x j)计算结果㊂K(x i,x j)=φ(x i)Tφ(x j),(6)由上所述,最终的超平面对应的模型为:f(x)=ðm i=1αi y i K(x i,x j)+b,(7)式中,K(x i,x j)为SVM模型训练时需要用到的核函数,也是用于实现从低维空间映射到高维空间的功能㊂为了构建适合于本文的行为识别模型,在构建分类器模型时先后选用了线性核函数㊁二次多项式核函数㊁三次多项式核函数以及径向基核函数训练模型,通过比较各模型的准确率选择最符合要求的核函数㊂4.2㊀多票判决由于超过1.28s后的数据会被划分到多个窗口之内,并且在每个窗口内被标记的标签类型可能不一致,造成该问题的原因有两种:①当前位置处于行为之间的交界处;②该窗口内的特征不明显,经分类器识别时误判㊂为了解决上述问题,本文参考多票判决策略㊂如图11所示,在分类器进行行为识别时,同时对每个时刻被标记的标签进行计数,直到当前时刻最后一次被划分到窗口内,选择被标记次数最多的标签作为此时刻的最终标签㊂若出现次数最多的标签有两个或两个以上,选择最后一个已知的标签作为最终标签㊂例如图11中标签7和9对应的行为分别为手机在位置2时的行走和上楼,其中某时刻的数据经过了6个窗口,被标记为9的次数最多,因此当前时刻最终预测出的行为为阅读时上楼㊂实验中,在线阶段采用固定长度为64,移动步长为10的窗口,因此,每个时刻最多经过6个窗口㊂图11㊀多票判决Fig.11㊀Multiple verdict5㊀实验结果与分析本文所有分类器模型均基于Matlab2020a编写代码,在Windows10系统,CPU为Intel i5-10210U 的电脑上执行代码㊂5.1㊀分类器实验结果分析在使用SVM训练分类器模型时,分别采用不同的核函数以及 一对一 的多分类方式,并采用5折交叉验证的方式避免过拟合㊂不同核函数的训练和测试结果如表7所示㊂表7㊀不同分类模型的训练和测试结果Tab.7㊀Training and testing results of differentclassification models核函数平均准确率/%训练时间/s预测时间/ms 线性核函数90.984.73230.88二次多项式核函数97.689.16232.42三次多项式核函数97.6173.25338.96径向基核函数95.2252.33280.38㊀㊀综合比较4种核函数的训练结果,其中采用二次多项式核函数的SVM模型(Q-SVM)整体性能最好㊂为探究SVM分类器模型性能的好坏,本文又分别训练并测试了朴素贝叶斯㊁决策树㊁KNN以及BP神经网络。
焦煤镜质组热解过程中胶质体的结构演化特性研究
㊀第36卷第2期煤㊀㊀质㊀㊀技㊀㊀术Vol.36㊀No.2㊀2021年3月COAL QUALITY TECHNOLOGYMar.2021移动阅读田鑫,王杰平,孙章.焦煤镜质组热解过程中胶质体的结构演化特性研究[J].煤质技术,2021,36(2):58-65.TIAN Xin,WANG Jieping,SUN Zhang.Structure evolutions of plastic phase derived from vitrinite in coking coal dur-ing pyrolysis process [J].Coal Quality Technology,2021,36(2):58-65.焦煤镜质组热解过程中胶质体的结构演化特性研究田㊀鑫,王杰平,孙㊀章(华北理工大学化学工程学院,河北唐山㊀063210)摘㊀要:研究炼焦煤热解成焦过程中胶质体阶段的结构对于认识成焦机理和指导配煤炼焦具有重要作用㊂采用基氏流动度测定仪对炼焦煤中镜质组进行热解实验以探究热解过程中胶质体的结构演化,并对特征温度点(T p ㊁T m ㊁T k )产物进行X 射线衍射(XRD )㊁拉曼光谱(Raman )㊁13C NMR 结构表征㊂结果表明:随着热解温度的增加,微晶结构参数中的堆垛高度(L c )㊁芳香度(f a )增加,石墨片层间距(d 002)㊁碳微晶尺寸(L a )减小;Raman 结构参数中的碳结构石墨化程度(A G /A all )增加,而热解产物中碳结构的缺陷结构和无序结构逐渐减小;核磁中芳香族碳含量逐渐增加,脂肪族含量减小㊂镜质组热解过程中胶质体流动度的变化分为胶质体发展过程与胶质体固化过程2个阶段,且2个阶段热解产物的结构演化速度存在明显差异,产物的结构参数在胶质体的发展过程中均演化较慢,而在胶质体的固化阶段演化较快㊂关键词:镜质组;结构演化;热解;胶质体;流动度;芳香度;X 射线衍射;拉曼光谱中图分类号:TQ520．1㊀㊀㊀文献标志码:A㊀㊀㊀文章编号:1007-7677(2021)02-058-08收稿日期:2021-01-10㊀㊀责任编辑:何毅聪㊀㊀DOI :10．3969/j．issn．1007-7677．2021．02．010㊀㊀基金项目:国家自然科学基金煤炭联合基金资助项目(U1361212)㊀㊀作者简介:田㊀鑫(1995 ),女,河北唐山人,华北理工大学在读硕士研究生㊂E -mail:894610598@qq．com㊀㊀通讯作者:孙㊀章(1979 ),男,安徽六安人,副教授㊁博士,主要从事煤化工的研究与教学㊂E -mail:sunz@ncst．edu．cnStructure evolutions of plastic phase derived from vitrinite in coking coalduring pyrolysis processTIAN Xin,WANG Jieping,SUN Zhang(College of Chemical Engineering ,North China University of Science and Technology ,Tangshan ㊀063210,China )Abstract :It is important to study the structures of the plastic phase (metaplast)derived from coking coals during py-rolysis and coking process for understanding the coking mechanism and guiding coal blending and coking.The pyroly-sis of the vitrinite in a coking coal was carried out by a Gieseler plastometer to investigate the structure evolutions of the metaplast during pyrolysis process,and the products at characteristic temperature (T p ,T m ,T k )were studied by X -ray diffraction (XRD),Raman spectroscopy (Raman),carbon -13nuclear magnetic resonance (13C NMR).The results show that with the increase of pyrolysis temperature,the stacking height (L c )and aromaticity (f a )of the metaplast increased,and the graphite interlayer spacing (d 002)and microcrystal size (L a )of the metaplast de-creased;the graphitizing degree of the carbon structure (A G /A all )for the metaplast increased,and the defective and disordered carbon structures of the metaplast gradually decreased;the aromatic carbon content in the metaplast in-creased,and the aliphatic content in the metaplast decreased.The fluidity of metaplast derived from the vitrinite dur-ing pyrolysis process was divided into two stages:the development and solidification of the metaplast.Furthermore,第2期田㊀鑫等:焦煤镜质组热解过程中胶质体的结构演化特性研究there was a great difference for the structure evolution rate of the metaplast between the two stages.The change rates for the structural parameters of the metaplast were slower in the development stage and faster in the solidification stage.Key words:vitrinite;structural evolution;pyrolysis;metaplast;fluidity;aromaticity;X-ray diffraction;Raman spectroscopy0㊀引㊀㊀言胶质体是炼焦煤在热解过程中形成的气㊁液㊁固三相混合物,是煤能够黏结成焦的重要阶段㊂胶质体的质量与数量直接决定所形成焦炭的性能,因此研究炼焦煤热解成焦过程中胶质体阶段的结构演化对于认识成焦机理和指导配煤炼焦具有重要作用㊂目前,许多学者[1-5]对胶质体的性质和结构进行了广泛的研究㊂Zubkova[4]用X射线衍射法研究了不同黏结能力煤中塑性层的结构,确定了塑性层结构对结焦机理的影响;杨野[5]研究了炼焦煤的挥发分析出与胶质体阶段的流动过程,发现两者相互作用决定了焦炭气孔结构的差异㊂陈鹏[6]探究了气煤在结焦过程中胶质体的变化,发现气煤胶质体质量差导致黏结性低㊂对于胶质体结构的表征,常采用XRD[7-9]㊁FTIR[10-11]㊁Raman[12-13]㊁TG-MS[14-16]㊁13C-NMR[17-18]等现代分析手段㊂Soonho L[19]和Chen[20]利用ATR-FTIR分析了不同煤种塑性层结构参数和官能团的变化㊂Hu[21]等利用XRD㊁SEM㊁FTIR等研究了焦煤在塑性层形成过程中的炭化特性及炭化行为㊂镜质组是形成塑性层的关键组分,许多学者[21-25]从显微组分角度开展热解行为研究,Tran[24]富集焦煤和非焦煤的显微组分,研究镜质组含量对煤热塑性和热解产物的影响;Wei[25]研究了挥发性焦油和胶质体萃取物等流动相对原煤热膨胀特性的影响㊂上述研究主要关注煤岩显微组分含量对热解产物的影响,而关于煤岩显微组分尤其是镜质组形成胶质体的微观结构演化的研究相对较少㊂因此,笔者采用基氏流动度测定仪对焦煤中镜质组进行热解实验,对胶质体特征温度点(T p㊁T m㊁T k)的产物进行XRD㊁FT-IR㊁13C NMR等结构分析,研究胶质体在流动度不同阶段结构的演化,为完善配煤炼焦机理提供实验基础和理论依据㊂1㊀实验部分1．1㊀镜质组的富集按照行业标准‘烟煤的镜质组密度离心分离方法“(MT/T807-1999)对斯凯拉斯焦煤的镜质组进行富集㊂首先将煤样粉碎至1．5mm以下,用1．40kg/L的ZnCl2密度液与煤混合均匀,利用离心机离心分离去除煤中的矿物质㊂烘干后再进行精细分离,将煤样粉碎至0．2mm以下,与1．30kg/L的ZnCl2密度液混合均匀,放入离心机离心分离,得到上层物为镜质组富集物㊂将镜质组富集物进行工业分析(水分:M ad;灰分:A d;挥发分:V daf)㊁黏结性测试(黏结指数:G)和煤岩分析(镜质组平均最大反射率:R max;镜质组含量:V t),所得结果见表1㊂表1㊀镜质组的基础性质指标Table1㊀Properties of vitrinite% M ad A d V daf G R max V t 0．512．2421．9291．71．40595．31．2㊀热解实验热解实验在基氏流动度测定仪(美国, PREIER4000)中进行㊂将已经安置好搅拌桨的甑坩埚置入装样器中,填入5g镜质组后压实㊂从装样器中取出坩埚,拧上坩埚盖,放入导向环,旋入基氏塑性仪头,使之下降至325ħ的金属浴中,以3ħ/min升温,实时观察流动度的变化,分别在流动度为1．00ddpm(软化温度T p)㊁最大流动度(最大流动温度T m)和流动度重新回到0．00ddpm(固化温度T k)时停止实验,冷却至室温,得到开始软化温度点热解产物(V-T p)㊁最大流动温度点热解产物(V-T m)㊁固化温度点热解产物(V-T k)㊂3个特征温度点(T p㊁T m㊁T k)的示意如图1所示㊂1．3㊀XRD测试为了探究实验样品微晶结构,采用日本理学会95煤㊀㊀质㊀㊀技㊀㊀术2021年第36卷图1㊀斯凯拉斯焦煤的流动度曲线及特征温度Fig．1㊀Gieseler fluidity development and characteristictemperatures of SKLS coking coal株式D /MAX2500PC 型X 射线衍射仪(40KV,150mA),以10ʎ/min 在10ʎ~60ʎ范围内进行扫描,其中入射波长为0．15418nm㊂煤焦的XRD 谱图有100峰与002峰2个特征峰㊂为了更好地分析结构变化规律以及计算微晶结构参数,需将002峰用软件进行分峰拟合[26],分峰拟合的示意如图2所示㊂分峰拟合处理后获得002峰和100峰的峰位和半峰宽,并通过谢乐公式以及布拉格方程[27]的计算得到石墨片层间距(d 002)㊁碳微晶尺寸(L a )㊁堆垛高度(L c )㊁芳香度(f a )等微晶结构参数㊂微晶结构参数的计算如下:L c =k 2λβ002cos θ002(1)d 002=λ2sin θ002(2)L a =k 1λβ100cos θ100(3)f a =A 002A γ+A 002(4)㊀㊀式中,l 为入射X 射线的波长;k 为与晶格形状及晶面指数有关的常数,k 1=1．84,k 2=0．94;Aγ㊁A 002分别为γ峰面积和002峰面积㊂1．4㊀Raman 测试为了探究实验样品的碳结构有序状态,采用美国赛默飞世尔公司DXR 型激光拉曼光谱仪进行测试,激光波长为532nm,分辨率为1cm -1,扫描范围为(100~3500)cm -1㊂Raman 测试过程中选取3个点,最终结果取3次结果平均值㊂煤焦的Raman 谱图有2个明显的峰[28],分别是1350cm -1左右的D 峰代表缺陷结构,1580cm -1左右的G 峰代表石墨结构㊂为了定量分析煤焦的结图2㊀热解产物V -T p 的002峰分峰拟合示意图Fig．2㊀The fitting diagram of 002peaks of V -T p构变化,对拉曼谱线进行分峰拟合处理,分峰拟合的示意如图3所示㊂D1㊁D2表示缺陷结构峰,D3属于无序性碳结构峰,D4表示交联结构峰,G 峰代表石墨结构峰,详见表2[29]㊂图3㊀热解产物V -T p 的Raman 峰及分峰拟合曲线Fig．3㊀The Raman spectrum and fitting curves of the V -T p char表2㊀Raman 拟合峰及其振动模式Table 2㊀The Raman bands and vibration modes峰位置/cm -1名称代表意义1330~1368D1晶粒边界的无序性和结构的缺陷1560~1620D2无序的石墨层片振动1475~1538D3有机分子㊁分子碎片㊁官能团等无定形的sp 2杂化碳原子振动~1208D4脂肪族结构或类烯烃结构中C C 的伸缩振动1587~1599G规则排列的sp 2碳原子伸缩振动形成的石墨峰1．5㊀固体13C -NMR 测试为了探究实验样品分子结构碳型,采用Bruker 400M 光谱仪测得样品的核磁共振光谱㊂实验采用4mm 样品转子,在双探头上进行㊂采用高分辨率固态核磁共振(NMR)谱,MAS 自旋速率为106第2期田㊀鑫等:焦煤镜质组热解过程中胶质体的结构演化特性研究Khz;回收时间:4s;用于采集的脉冲程序:cp;预扫描延迟:6．5μs㊂2㊀结果与讨论2．1㊀镜质组的基氏流动度斯凯拉斯煤镜质组富集物的基氏流动度特征曲线如图4所示㊂从图4中看出热解温度上升到429ħ(T p )时流动度达到1．00ddpm,说明此时镜质组开始软化热解;温度继续上升,镜质组继续热解,流动度逐渐增大,到471ħ(T m )时达到峰值,随后流动度随着热解温度的上升开始逐渐下降,到493ħ(T k )时回到0．0ddpm,说明此时镜质组已经固化,不再具有流动性㊂图4㊀镜质组的基氏流动度曲线及特征温度Fig．4㊀Gieseler fluidity development and characteristictemperatures of vitrinite由图4中流动度变化的发展趋势来看,镜质组的胶质体演化可以分为2个阶段,从开始软化至最大流动度为第1阶段(T p -T m ),此阶段内流动度的变化速率较为缓慢,经历14min (1．70ddpm /ħ),温度间隔42ħ;从最大流动度至胶质体固化为第2阶段(T m -T k ),此阶段内流动度的变化速率较快,经历7．3min (-3．29ddpm /ħ),温度间隔22ħ㊂前后两个阶段的流动度变化速率和温度区间的差异说明胶质体的相态随着温度的升高非均匀变化,流动度最大值时是胶质体相态的转变点㊂前一阶段是胶质体发展过程,镜质组缓慢热解软化,煤粒间相互融并生成胶质体,流动度逐渐增大,后一阶段为胶质体的固化过程,镜质组的热解速度较快,胶质体的自由基相互结合形成大分子片层,逐渐固化,所以流动度迅速减小[30]㊂2．2㊀XRD 分析镜质组富集物在3个特征温度点热解产物的XRD 如图5所示㊂由图5可知,3个热解产物均表现出较明显的002峰,而100峰不明显㊂且随着热解温度的升高,002峰逐渐变得尖锐,而100峰变化不太大,说明热解产物的芳香层片尺寸逐渐变化㊂图5㊀镜质组特征热解产物的XRD 谱图Fig．5㊀The XRD profiles for pyrolysis products of vitrinite对3种热解产物的XRD 进行分峰拟合得出的微晶结构参数见表3㊂由表3可知,随着胶质体的演化,热解产物的堆叠高度(L c )逐渐增加,层片间距(d 002),碳微晶尺寸(L a )逐渐减少㊂此为镜质组热解析出的小分子起到润滑作用促进芳香层片在纵向的堆叠,从而导致L c 变大㊂随着热解的进行,挥发分不断析出,破坏了部分交联键,使芳香结构更加紧密,d 002和L a 逐渐变小㊂芳香度(f a )在热解过程中逐渐变大,说明非晶型的碳正在向晶型碳转变[31]㊂表3㊀镜质组特征热解产物的微晶结构参数Table 3㊀Microcrystal structural parameters for pyrolysisproducts of vitriniteChar L c /nm d 002/nm L a /nm f a V -T p2．2080．3531．7450．735V -T m 2．2360．3511．4760．748V -T k 2．2660．3501．2480．757㊀㊀为了进一步分析胶质体结构的演化特性,微晶结构参数随热解温度变化趋势如图6所示㊂由图6可看出,在热解前期的胶质体发展阶段,胶质体的堆叠高度(L c )和碳微晶尺寸(L a )的变化速度较慢,而后期的胶质体固化阶段,堆叠高度(L c )和碳微晶尺寸(L a )变化速度较快㊂因为前一阶段的胶质体生成过程,主要是镜质组的大分子解聚㊁脂肪侧链的断裂和小分子的析出,碳微晶的转化速度较慢;后一阶段的胶质体固化过程,主要为大量芳香自由基缩聚增大,石墨晶格发展速度较16煤㊀㊀质㊀㊀技㊀㊀术2021年第36卷快㊂层片间距(d 002)前期变化较快而后期较慢的原因是主要是前期小分子析出较多促进层片间距增加,而后期小分子析出较少,则层片间距增幅变小[32]㊂热解产物的芳香度(f a )前期两阶段的变化趋势相差不大,说明胶质体的相态在热解转变过程中的碳微晶逐渐增加,后期稍有增加㊂图6㊀镜质组特征热解产物的结构参数与热解温度的关系Fig．6㊀Relationships of the structural parameters for vitrinite pyrolysis products with temperature2．3㊀Raman 分析镜质组3个特征热解产物所测得的Raman 谱图如图7所示㊂由图7可看出,3个热解产物均具有2个特征峰,即1350cm -1的缺陷峰(D 峰)和1580cm -1的石墨峰(G 峰)㊂且随着热解温度的升高,G 峰明显变得尖锐,说明胶质体中石墨微晶结构在逐渐增加㊂图7㊀镜质组特征热解产物的Raman 谱图Fig．7㊀The Raman profiles for pyrolysis products of vitrinite对3种热解产物的Raman 谱图进行分峰拟合得出的结构参数见表4㊂从表4中可知,随着胶质体的演化,热解产物中碳结构的缺陷和无序结构(A D1/A G ㊁A D2/A G ㊁A D3/A G ㊁A D4/A G )逐渐减小,碳结构的石墨化程度(A G /A all )逐渐变大㊂因随着热解的进行,胶质体中大分子发生解聚-缩聚反应,缺陷结构与无序结构减少,促使无序与缺陷的碳结构转化为石墨化碳结构[33]㊂表4㊀镜质组特征热解产物的Raman 结构参数Table 4㊀Raman structural parameters for pyrolysisproducts of vitriniteChar A D1/A G A D2/A G A D3/A G A D4/A G A G /A all V -T p2．4790．3030．4200．7520．202V -T m 2．1910．2320．3940．6840．222V -T k 1．9080．1640．3660．6390．245㊀㊀为了进一步分析胶质体碳结构的演化特性,Raman 结构参数随热解温度变化趋势如图8所示㊂由图8可看出,缺陷与无序碳结构(A D1/A G ㊁A D2/A G ㊁A D3/A G ㊁A D4/A G )和石墨化碳结构(A G /A all )均在热解前期的胶质体发展阶段的变化速度较慢,而后期的胶质体固化阶段的变化速度较快(近似2倍)㊂因为前一阶段是胶质体的生成阶段,主要镜质组中芳香基团侧链的裂解反应为主,产生大量挥发性小分子,相邻的芳香自由基相互聚合增大,所以缺陷结构与无序结构缓慢减小,石墨化碳结构缓慢增加㊂而后一阶段是胶质体的固化阶段,大量生成的芳香自由基相互聚合增大转变为固体,无序结构与缺陷碳结构快速的转化为石墨化碳结构㊂2．4㊀13C NMR 分析镜质组富集物在3个特征温度点热解产物的核磁谱图如图9所示㊂位于(90~170)ˑ10-6的芳香碳的峰强度要明显高于脂肪族碳,说明热解产物中的碳主要以芳香碳为主,脂肪碳含量较少㊂对3种热解产物13C NMR 谱图进行积分,根据面积计算所得的各类碳含量见表5㊂26第2期田㊀鑫等:焦煤镜质组热解过程中胶质体的结构演化特性研究图8㊀镜质组特征热解产物的Raman 结构参数与热解温度的关系Fig．8㊀Relationship of Raman structural parameters for vitrinite pyrolysis products withtemperature图9㊀镜质组特征热解产物的核磁谱图Fig．9㊀The 13C NMR profiles for pyrolysis products of vitrinite表5㊀镜质组特征热解产物的13C NMR 结构参数Table 5㊀13C NMR structural parameters for pyrolysisproducts of vitriniteChar f a f al V -T p0．1960．804V -T m 0．1760．824V -T k 0．1370．863㊀㊀由表5可知,随着胶质体相态的演化,热解产物的脂肪碳含量(f al )逐渐减小,芳香碳含量(f a )逐渐变大㊂因为随着镜质组的热解,胶质体中脂肪侧链不断脱落,芳香自由基不断聚合,脂肪碳分解成小分子逸出,导致脂肪族碳含量减少㊁芳香碳含量增加[34]㊂为进一步分析胶质体中碳类型的演化特性,13CNMR 结构参数随热解温度变化趋势如图10所示㊂由图10可知,脂肪碳和芳香碳的含量在热解前期的胶质体发展阶段的变化速度较慢,而后期的胶质体固化阶段的变化速度较快㊂由此说明镜质组生成的胶质体前期主要以大分子的断键反应为主,后期主要是芳香自由基的聚合增大反应为主[35]㊂图10㊀镜质组特征热解产物核磁结构参数与热解温度的关系Fig．10㊀Relationship between structural parameters of vitrinitepyrolysis products and pyrolysis temperature3㊀结㊀㊀论(1)斯凯拉斯焦煤镜质组热解过程中胶质体按流动度的变化分为缓慢增加的发展过程(T p -T m )和快速减小的固化过程(T m -T k )2个阶段㊂随热解温度的升高,胶质体特征产物的XRD 微晶36煤㊀㊀质㊀㊀技㊀㊀术2021年第36卷结构参数L c㊁f a逐渐增加,d002与L a逐渐减小㊂(2)Raman结构参数中除石墨结构参数A G/A all增加外,其余参数A D1/A G㊁A D2/A G㊁A D3/A G㊁A D4/A G不断减小;脂肪碳含量不断减小,芳香碳含量不断增加㊂(3)结构参数在胶质体发展过程与固化过程2个阶段的结构演化程度存在明显差异㊂XRD微晶结构参数,Raman结构参数㊁核磁结构参数在胶质体的发展过程演化较慢,而在胶质体的固化阶段演化较快㊂综上所述,胶质体的流动度变化速率与胶质体的微观结构参数具有一定的相关性,能够反映胶质体气-固-液三相混合物的相态演化特性㊂参考文献(References):[1]㊀陈鹏,薛改凤,贾丽晖,等.炼焦煤水分对胶质层质量的影响研究[J].武钢技术,2016,54(3):4-6.[2]㊀TRIPATHY A,BISWAL S K,MEIKAP B C.Statisti-cal modelling and optimization study for beneficiation ofIndian high ash semi-coking coal using allflux separator[J].Advanced Powder Technology,2016,27(4):1488-1493.[3]㊀赵奇.吉氏流动度在煤炭焦化领域的应用[J].洁净煤技术,2017,23(6):60-64.ZHAO Qi.Application of Gieseler fluidity on coal co-king field[J].Clean Coal Technology,2017,23(6):60-64.[4]㊀ZUBKOVA V V.Some aspects of structural transforma-tions taking place in organic mass of Ukrainiancoals during heating.Part1.Study of structural trans-formations when heating coals of different caking capaci-ty[J].Fuel,2005,84(6):741-754.[5]㊀杨野,付铁林,赵宇科,等.炼焦煤热解过程交互行为与气孔特征的关系[J].燃料与化工,2020,51(4):24-26.YANG Ye,FU Tielin,ZHAO Yuke,et al.The rela-tion between the interaction behavior of coking coal py-rolysis and pore characteristics[J].Fuel&ChemicalProcesses,2020,51(4):24-26.[6]㊀陈鹏,张雪红,项茹.气煤结焦过程行为浅析[J].山东化工,2015,44(23):20-21.CHEN Peng,ZHANG Xuehong,XIANG Ru.Analy-sis of the Behavior of Coal Coking Process[J].Shan-dong Chemical Industry,2015,44(23):20-21.[7]㊀LU L,SAHAJWALLA V,HARRIS D.Characteristicsof Chars Prepared from Various Pulverized Coals at Dif-ferent Temperatures Using Drop-Tube Furnace[J].Energy Fuels,2000,14(4):869-876.[8]㊀YAN J,LEI Z,LI Z,et al.Molecular structure char-acterization of low-medium rank coals viaXRD,solidstate13C NMR and FTIR spectroscopy[J].Fuel,2020,268:117038.[9]㊀刘冬冬,高继慧,吴少华,等.热解过程煤焦微观结构变化的XRD和Raman表征[J].哈尔滨工业大学学报,2016,48(7):39-45.LIU Dongdong,GAO Jihui,WU Shaohua,et al.XRD and Raman characterization of microstructure chan-ges of char during pyrolysis[J].Journal of Harbin In-stitute of Technology,2016,48(7):39-45. [10]㊀ZHANG X,ZHOU B,AN D,et al.Effect of heatingrate on pyrolysis characteristics and char structure ofZhundong lignite coal[J].Journal of China Coal So-ciety,2019,44(2):604-610.[11]㊀YAN Z,QIU P,GANG C,et al.Selective enrich-ment of chemical structure during first grinding ofZhundong coal and its effect on pyrolysis reactivity[J].Fuel,2017,189:46-56.[12]㊀PAN J,LV M,BAI H,et al.Effects of metamor-phism and deformation on the coal macromolecularstructure by laser Raman spectroscopy[J].Energy&Fuels,2017,31(2):1136-1146.[13]㊀ZHONG M,ZHAO Y,ZHAI J R,et al.Effectsof nickel additives with different anions on the struc-ture and pyrolysis behavior of Hefeng coal[J].FuelProcessing Technology,2019,193:273-281. [14]㊀ZHANG Y,LI Z.Raman spectroscopic studyof chemical structure and thermal maturity of vitrinitefrom a suite of Australia coals[J].Fuel,2019,241:188-198.[15]㊀XU S,ZENG X,HAN Z,et al.Quick pyrolysis ofa massive coal sample via rapid infrared heating[J].Applied Energy,2019,242:732-740.[16]㊀王小华,赵洪宇,李玉环,等.基于TG-MS研究不同升温速率下褐煤热解气体产物析出特性及其动力学参数[J].煤炭工程,2018,50(3):140-144.WANG Xiaohua,ZHAO Hongyu,LI Yuhan,etal.Study on releasing characterristics of pyrolysis gasproducts and kinetic analysis of lignite pyrolysis at dif-ferent heating rates based on TG-MS[J].Coal Engi-neering,2018,50(3):140-144.[17]㊀KAWASHIMA H,YAMASHITA Y,SAITO I.Stud-ies on structural changes of coal density-separa-ted components during pyrolysis by means of solid-state13C NMR spectra[J].Journal of Analytical andApplied Pyrolysis,2000,53(1):35-50.46第2期田㊀鑫等:焦煤镜质组热解过程中胶质体的结构演化特性研究[18]㊀BAYSAL,MUSTUFA,YURUM,et al.Structure ofsome western Anatolia coals investigated by FTIR,Ra-man,13C solid state NMR spectroscopy and X-ray dif-fraction[J].International Journal of Coal Geology,2016,163:166-176.[19]㊀SOONHO L,YU J,MERRICK M,et al.A studyon the structural transition in the plastic layer dur-ing coking of Australian coking coals using Synchro-tron micro-CT and ATR-FTIR[J].Fuel,2018,233:877-884.[20]㊀CHEN Y,LEE S,TAHMASEBI A,et al.Chemicalstructure transformation during the later stage of plasticlayers during coking using Synchrotron infrared mi-crospectroscopy technique[J].Fuel,2020,273:117764.[21]㊀HU W J,WANG Q,ZHAO X F,et al.Relevancebetween various phenomena during coking coal carboni-zation.Part2:Phenomenon occurring in the plasticlayer formed during carbonization of a coking coal[J].Fuel,2019,253:199-208.[22]㊀CHANG H,DENG H,YANG Q,et al.Interactionof vitrinite and inertinite of Bulianta coal in pyrolysis[J].Fuel,2017,207:643-649.[23]㊀LI M,YANG J H,ZHANG Q F,et al.Struc-ture change and organic sulfur forms transforma-tion during pyrolysis of high-sulfur vitrinite[J].Journal of Fuel Chemistry&Technology,2014,42(2):138-145.[24]㊀TRAN Q A,STANGER R,XIE W,et al.Thermo-plastic Development of Coking and Non-Coking Macer-al Concentrates and Molecular Weight Distribution ofTheir Pyrolysis Products[J].Journal of Analytical&Applied Pyrolysis,2017:72-85.[25]㊀WEI X,STANGER R J,TRAN Q A,et al.Impactof Coal Pyrolysis Products as a Rheological Additive onThermoplasticity of a Coking Coal[J].Energy&Fu-els,2018,32(4):15-35.[26]㊀王英新界,王重阳,张超,等.艾维尔沟煤中镜质组配煤在成焦过程中的结构演化[J].煤质技术,2020,35(1):6-11.WANG Yingxinjie,WANG Chongyang,ZHANG Chaoet al.Structure evolutions of chars derived from Ewir-gol vitrinite blending coals during coking process[J].Coal Quality Technology,2020,35(1):6-11.[27]㊀WU D,ZHNG W.Evolution Mechanism of Macromo-lecular Structure in Coal during Heat Treatment:Based on FTIR and XRD In Situ Analysis Techniques[J].Journal of Spectroscopy,2019:1-18.[28]㊀李金泽,王杰平,孙章.煤及其热解过程中微观结构的光谱学研究进展[J].燃料与化工,2020,51(2):8-12.LI Jinze,WANG Jieping,SUN Zhang.Progressof microstructures of coals and those during pyrolysisprocess by spectroscopy[J].Fuel&ChemicalProcesses,2020,51(2):8-12.[29]㊀XIE Y,YOU J,LU L,et al.Raman Spectroscop-ic Study of Coal Samples during Heating[J].Ap-plied Sciences,2019,9(21):4699.[30]㊀王春花.煤的基氏流动度研究进展[J].梅山科技,2008(1):5-7.WANG Chuanhua.Method for coal blending co-king through utilizing Gieseler fluidity index[J].Baosteel Meishan,2008(1):5-7.[31]㊀张金刚,孙志刚,郭强,等.神府煤热解的结构变化及煤焦加氢反应性研究[J].燃料化学学报,2017,45(2):129-137.ZHANG Jingang,SUN Zhigang,GUO Qiang,etal.Structural changes of Shenfu coal in pyrolysisand hydrogasification reactivity of the char[J].Jour-nal of Fuel Chemistry and Technology,2017,45(2):129-137.[32]㊀张小蕊,邹冲,赵俊学,等.XRD和Raman法评估热解气氛中H2和CO对半焦化学结构的影响[J].燃料化学学报,2019,47(11):1288-1297.ZHANG Xiaorui,ZOU Chong,ZHAO Junxue,et al.Effect of H2and CO as pyrolysis atmosphere on chemi-cal structure of char by XRD and Raman methods[J].Journal of Fuel Chemistry and Technology,2019,47(11):1288-1297.[33]㊀ZHU H,YU G,GUO Q,et al.In Situ Raman Spec-troscopy Study on Catalytic Pyrolysis of a BituminousCoal[J].Energy&Fuels,2017,31(6):5817-5827.[34]㊀王擎,石聚欣,迟铭书,等.基于13C NMR技术的桦甸油页岩热解行为[J].化工进展,2014,33(9):2321-2325.WANG Qing,SHI Juxin,CHI Mingshu,et al.Pyroly-sis behavior of Huadian oil shale based on13CNMR technology[J].Chemical Industry and Engi-neering Progress,2014,33(9):2321-2325. [35]㊀KAWASHIMA H,YAMASHITA Y,SAITO I.Stud-ies on structural changes of coal density-separa-ted components during pyrolysis by means of solid-state13C NMR spectra[J].Journal of Analytical andApplied Pyrolysis,2000,53(1):35-50.56。
有机小分子过氧化氢荧光探针研究进展及发展方向与应用前景
基金项目：四川省科技计划项目(2018JY0168)通信作者：张凯，E-mail：***************.cn 引用本文：张燕军，李洁丽，张凯.有机小分子过氧化氢荧光探针研究进展及发展方向与应用前景[J].西南医科大学报.2023，46(6)：541-546.DOI ：10.3969/j.issn.2096-3351.2023.06.016活性氧（reactive oxygen species ，ROS ）是一类高活性的含氧物质，在细胞的增殖、生长、免疫、信号传递等过程发挥着重要的作用[1]。
过氧化氢（hydrogen perox‑ide ，H 2O 2）是人体内ROS 的重要组成部分，它主要来源于细胞线粒体，经线粒体传递链（electron transport chain ，ETC ）复合体等途径产生[2-3]。
H 2O 2对细胞具有双重的生理作用：适量的H 2O 2有利于细胞的生长、分化和维持[4-5]；而过量的H 2O 2会引起细胞的损伤、凋亡和自噬，从而诱发炎症、心脑血管疾病、老年痴呆（senile dementia ，SD ）和癌症等多种疾病[6-8]。
因此，精确识别和检测细胞内H 2O 2对生命科学研究和临床诊断都具有非常重要的意义。
针对H 2O 2常见的检测方法主要有：比色法[9]、电化学法[10]、质谱分析法[11]和荧光分析法[12]等。
其中，基于小分子荧光探针的荧光分析法由于具有操作简单、生物相容度高、灵敏度高、选择性强、可实时原位检测等特点而备受关注；将荧光分析法与激光共聚焦显微镜技术结合而产生的荧光成像技术，也成为生物分析、检测的重要手段，应用于生物学研究、疾病早期诊断和临床治疗评价等领域[13-14]。
近年来针对H 2O 2荧光检测和成像的研究已成为有机小分子探针发展的重要方向，本文将从小分子荧光探针的响应机理进行归纳总结，介绍H 2O 2荧光探针的设计、构建及生物应用，并展望其发展方向和应用前景。
ru单原子-氮化碳界面调控协同压电效应强化光催化生物质氧化耦合产氢研究
ru单原子-氮化碳界面调控协同压电效应强化光催化生物质氧化耦合产氢研究下载温馨提示：该文档是我店铺精心编制而成，希望大家下载以后，能够帮助大家解决实际的问题。
文档下载后可定制随意修改，请根据实际需要进行相应的调整和使用，谢谢!并且，本店铺为大家提供各种各样类型的实用资料，如教育随笔、日记赏析、句子摘抄、古诗大全、经典美文、话题作文、工作总结、词语解析、文案摘录、其他资料等等，如想了解不同资料格式和写法，敬请关注!Download tips: This document is carefully compiled by the editor. I hope that after you download them, they can help you solve practical problems. The document can be customized and modified after downloading, please adjust and use it according to actual needs, thank you!In addition, our shop provides you with various types of practical materials, such as educational essays, diary appreciation, sentence excerpts, ancient poems, classic articles, topic composition, work summary, word parsing, copy excerpts, other materials and so on, want to know different data formats and writing methods, please pay attention!光催化技术作为一种环境友好的能源转化方式，在水裂解产氢和有机底物氧化等领域具有巨大的应用潜力。
《载体对Pt-W-Zr系催化剂催化氢解甘油性能的影响》范文
《载体对Pt-W-Zr系催化剂催化氢解甘油性能的影响》篇一一、引言随着人类对可再生能源和绿色化学的日益关注，生物质能源的开发和利用成为当前研究的热点。
其中，甘油作为生物柴油产业的副产品，其高效利用具有重要的经济和环境价值。
氢解甘油是一种有效的转化途径，能够得到多种高附加值的化学品。
在氢解甘油的过程中，催化剂的选择至关重要。
本文着重探讨载体对Pt-W-Zr系催化剂催化氢解甘油性能的影响。
二、Pt-W-Zr系催化剂的概述Pt-W-Zr系催化剂是一种广泛应用于氢解甘油反应的催化剂。
该催化剂体系中的Pt作为活性组分，W和Zr则作为助剂和载体元素。
催化剂的性能受其组成、结构以及制备方法等多方面因素的影响。
其中，载体的选择对于催化剂的性能具有显著的影响。
三、载体的种类及其作用载体在催化剂中起着重要的作用，它不仅提供了催化剂活性组分的支撑，还影响着催化剂的分散度、比表面积、孔结构以及催化性能等。
在Pt-W-Zr系催化剂中，常用的载体包括氧化铝、二氧化硅、氧化钛、活性炭等。
四、不同载体对Pt-W-Zr系催化剂催化氢解甘油性能的影响1. 氧化铝载体：氧化铝载体具有较高的比表面积和良好的热稳定性，能够提高催化剂的分散度和稳定性。
然而，过高的比表面积可能导致催化剂活性过高，从而增加副反应的发生。
2. 二氧化硅载体：二氧化硅载体具有较好的化学稳定性和机械强度，能够提高催化剂的抗毒性和耐久性。
此外，二氧化硅载体的表面性质可以通过改性来调节，从而优化催化剂的活性。
3. 氧化钛载体：氧化钛载体具有较高的表面能和较强的酸性，能够提高催化剂的活性。
然而，氧化钛载体的热稳定性较差，可能影响催化剂的长期性能。
4. 活性炭载体：活性炭载体具有较大的比表面积和良好的吸附性能，能够提高催化剂的分散度和反应物的吸附能力。
此外，活性炭载体的孔结构可以调节，有利于反应物的扩散和传输。
五、实验结果与讨论通过对比不同载体对Pt-W-Zr系催化剂催化氢解甘油性能的影响，我们发现：1. 使用氧化铝载体的催化剂在反应初期表现出较高的活性，但长期运行过程中易出现失活现象。
Si掺杂InN的电子结构和光学性质的第一性原理计算
2. 2 态密度分析 图 3是不同比例 Si掺杂 InN 的 总态密度变化对比情况. 由图可见 , Si掺杂后总态 密 度 下 价 带 的 峰 值 变 小. Si 掺 杂 比 例 为 0%、 6. 25%、12. 5%和 25%的下价带的最高峰分别位于
- 13. 22、- 14. 98、- 15. 16 和 - 15. 44 eV. 从总态 密度的变化能够看出 , Si掺杂后 InN 的电子趋于占 据低能级 ,而且掺杂比例越高趋势越明显 ,这与能 带结构所反映的趋势一致. 来自 In (4d)和 N (2 s)轨
1 理论模型和计算方法
计算采用基于密度泛函理论框架下的第一性 原理平面波超软赝势 ( PW P)方法 ,电子的交换关联 能用广义梯度近似 ( GGA )描述 [ 10211 ]. 构建的模型为 32原子的 2 ×2 ×2的纤锌矿结构超原胞 (如图 1) , 在各个方向上的原子数都是基本单原胞原子个数 的两倍 ,晶格常数采用实验数据 a = 0. 353 nm , c = 0. 569 nm , u = 0. 038 nm[ 12 ] ,属于 P63mc空间群. 在 超原胞中掺入不同比例的 Si,并取代 In的位置 ,但 并不改变晶体结构 ,原胞建立起来之后我们对其进 行优化. 最后 ,基于优化好的超晶胞 ,我们计算了它 们的带隙 、能带结构以及光学性质. 参与计算的价 态电子为 : N ( 2 s2 2p3 ) 、Si ( 3 s2 3p2 ) 、In ( 4d10 5 s2 5p1 ) ,
收稿日期 : 2008 - 05 - 01 基金项目 :四川省应用基础研究基金 (2006J132052)资助项目 3 联系作者简介 :徐 明 (19692) ,男 ,教授 ,主要从事凝聚态物理的研究
- 1、下载文档前请自行甄别文档内容的完整性，平台不提供额外的编辑、内容补充、找答案等附加服务。
- 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
- 3、如文档侵犯您的权益，请联系客服反馈,我们会尽快为您处理(人工客服工作时间：9:00-18:30)。
The Knowledge Engineering Review,V ol.00:0,1–4.c 2005,Cambridge University PressDOI:10.1017/S000000000000000Printed in the United Kingdom Representation in case-based reasoningRALPH BERGMANN1,JANET KOLODNER2and ENRIC PLAZA31University of Trier,Business Information Systems II,54286Trier,GermanyE-mail:bergmann@uni-trier.de2Georgia Institute of Technology,Atlanta,Georgia30032-0280,USAE-mail:janet.kolodner@3IIIA-Institut d’Investigaci´o en Intellig`e ncia Artificial,Campus UAB,08193Bellaterra,Catalonia,SpainE-mail:enric@iiia.csic.esAbstractA case in case-based reasoning is a contextualized piece of experience,which can be represented in various forms.Traditional approaches can be classified into three main categories:feature vector representations, structured representations,and textual representations.More sophisticated approaches make use of hierarchical representations or generalized cases.For particular tasks such as design and planning highly specific representations have been developed.1IntroductionOriginally,the notion of case was that of a“problem solving episode”,based on the cognitive science distinction between semantic and episodic memory.According to(Kolodner,1993)a case is a“contextu-alized piece of knowledge representing an experience that teaches a lesson fundamental to achieving the goals of the reasoner”.The experience a case represents can be structured in various ways.Very often it is only subdivided into a problem and a solution description,but additional knowledge might be necessary depending on the kind of intended reuse.For example,(Kolodner,1993)proposes a comprehensive case structure consisting of the followingfive parts:(i)a situation and its goal,(ii)the solution and,sometimes, means of deriving it,(iii)the result of carrying it out,(iv)explanations of results,and(v)lessons that can be learned from the experience.Case representation in case-based reasoning(CBR)makes use of familiar knowledge representation formalisms from AI to represent the experience contained in the cases for reasoning purposes.A large variety of representation formalisms have been proposed.However,three major types of case representation have arisen:feature vector(or propositional)cases,structured(or relational)cases,and textual(or semi-structured)cases.In addition,specialized representations of cases for specific tasks have been used,such as plans used as cases in planning.Feature-vector approaches represent a case as a vector of attribute-value pairs,similar to the propo-sitional representations used in Machine Learning(ML),that support k-nearest neighbor matching and instance-based learning(Aha et al.,1991).The structured approach to case representation originates from the episodic memory notion of cases.The representational structure itself is usually developed around a frame-based formalism.Since frames can be seen as a subset offirst-order logic,cases represented as frames or some frame-like structure are examples of what ML calls relational representations.Such representations arise from gathering together clusters of relations that occur together in elementary objects.Cases,from this perspective,are clusters of relations between the kinds of elementary objects that comprise them.Frame representations also strongly resemble object-oriented case representations (Bergmann,2002,Chapter3),that have a different origin and are often used in industrial applications. Finally,textual case representations(Lenz&Burkhard,1996)go in the opposite direction by imposing only a weak structure on the cases.This allows easy exploitation of the experience captured in documents2R.BERGMANN,J.KOLODNER AND E.PLAZAsuch as bug reports or FAQ’s.Textfields are usually represented as sets of linguistic items(e.g. words reduced by some stemming algorithm)similar to vector representations in information retrieval. However,the available(weak)structure of the text allows the introduction of more semantics for a better interpretation of the reusability of a case.Representation of cases and the way similarity is assessed during retrieval are strongly related to each other.Distance-based similarity metrics are easy to apply to feature vectors,while techniques related to Information Retrieval(IR)can easily be applied to textual representations.Frame-based cases often require knowledge-intensive indexing and matching algorithms.2Basic approaches to case representation2.1Content of a caseIn a representative paper on case representation,“Improving human decision making through case-based decision aiding,”Janet Kolodner(1991)focuses on the role cases can play in helping people make decisions and on the content cases need to have to play that role.Kolodner does not try to comment on the form that a case should take but rather focuses on the kinds of things that should be represented in a case so that it can be productively used for reasoning.Kolodner recommends that cases include a problem situation description,the solution that was proposed(often including how that solution was derived),and the outcome,including the state of the world after the solution was carried out,how close that was to what was expected,and explanations,if necessary and available,of why it might not have worked as well as expected.2.2Feature vector representationThe PROTOS system(Porter et al.,1990)uses a feature vector approach for domains with weak or intractable theories.A category is extensionally represented as a collection of cases called exemplars.A new case is classified into a category if a match can be found between an exemplar and the new case.This matching process is knowledge intensive and tries to build an explanation that connects the features of the new case with an exemplar.Since each explanation is a path constructed inside a semantic net,retrieval is the process of explaining the(similarity)relation between a new case and an exemplar.Unlike most early CBR systems that use feature vector representations,PROTOS already uses a knowledge-intensive similarity measure.2.3Frame-based representationFrame based representations have been(partially)formalized by description logics.The notion of“cases as terms”(Plaza,1995)argues that viewing structured cases as terms in feature logics(a particular brand of description logics)helps to better understand several aspects of case-based reasoning.Domain knowledge can be integrated using a sort hierarchy and the issue of composite cases(cases that group together other objects or sub-cases)is understood by the fact that a sub-term is also a term.Finally,the notion of similarity between two cases is linked to the concepts of subsumption and anti-unification of terms.This notion also bridges relational cases in CBR with relational learning in inductive ML where subsumption and anti-unification are basic building blocks.2.4Object-oriented representationObject-oriented case representations(Bergmann,2002,Chapter3)have an expressiveness similar to frame representations,but have a different origin.They make use of the data modeling approach of the object-oriented paradigm,including is-a and part-of relations as well as the inheritance principle.Cases are represented as collections of objects,each of which is described by a set of attribute-value pairs. The structure of an object is described by an object class.Several object-oriented case representation languages have been developed.CASUEL(Manago et al.,1994)is an early example in plain ASCII,while more recent languages are XML compatible.Object-oriented representations are particularly suitable forRepresentation in case-based reasoning3 complex domains in which cases with different structures occur.An impressive example is described by G¨o ker&Roth-Berghofer(1999).2.5Textual representationTextual case representations,as described by Lenz&Burkhard(1996),decompose the text that constitutes a case into information entities(IEs).An IE is a word or a phrase contained in the text that is relevant to determine the reusability of the episode captured in the case.Given a vocabulary of the relevant words or phrases,text cases can be mined for IEs,allowing case acquisition to be automated.The set of cases that form the case base is organized in the form of a case retrieval net(CRN),which is a directed graph with nodes representing cases and their IEs.These nodes are linked according to their similarity.Hence knowledge about similarity is encoded into the strength of the links between the nodes in the CRN. Case retrieval is similar to activation propagation in a neural network:the IEs that occur in the current problem are activated and this initial activation is propagated through the case retrieval net according to the similarity-based link strength.3Advanced approaches3.1Hierarchical case representationThe previously discussed approaches typically represent cases at a single level of abstraction.However, in recent CBR publications,the use of multiple representations at different levels of abstraction has been investigated(Bergmann&Wilke,1996).The basic idea behind these approaches is to represent a case at multiple levels of detail,possibly using multiple vocabularies.When a new problem must be solved, similar cases at appropriate levels of abstraction are retrieved from the case base and the solutions from these cases are combined and refined.JULIA(Hinrichs,1992;Kolodner,1993)provides one such hybrid representation for cases.It begins with a hierarchical frame-based representation that distinguishes between different parts of a complex case.Connecting the parts of a case,constraints show how they are related to each other.An advantage of this representation is that it allows a whole case or its parts to be accessed and used by the case-based reasoner,and the constraints can be used to guide adaptation.3.2Generalized casesIn the previous discussion,a case is regarded as a single experience,e.g.representing a point in the case space.In contrast,a generalized case covers a subspace of the representation space.A single generalized case immediately provides solutions to a set of closely related problems rather than to a single problem only.Kolodner(1983),Zito-Wolf&Alterman(1992),and Bergmann&V ollrath(1999)discuss this issue from different perspectives.In Zito-Wolf&Alterman’s work,for example,a single case contains a variety of alternative plans that are available for achieving a common goal.From a technical point of view, generalized cases are represented by introducing variables in a traditional case representation approach. Allowed assignments to these variables can be further restricted by introducing constraints in the case representation.3.3Cases in case-based designSpecific case representations have been developed for particular tasks like planning or design.In design, for instance,the DRAMA system(Leake&Wilson,1999)uses CMaps(Concept Maps)to represent aircraft design cases.A CMap is a semi-structured case containing both a structural aspect(similar to semantic nets to describe aircraft components)and a textual aspect(to document design decisions and rationale).The FABEL project developed CBR support systems for architectfirms in the design of buildings(Gebhardt et al.,1997).The representation of cases is based on CAD objects,which are described by coordinates,aspect,granularity,and size.On top of these objects,a FABEL case consists of a problem,a plan(a solution),and a collection of solution paths or explanations.The complexity of this4R.BERGMANN,J.KOLODNER AND E.PLAZAspecialized case representation gave rise to new or adapted techniques for retrieval and reuse(Gebhardt et al.,1997).Hinrichs’hierarchical representations in JULIA(Hinrichs,1992;Kolodner,1993)were also developed to support case-based design.3.4Cases in case-based planningCase representations for case-based planning(see Cox et al.this issue)are sometimes more specialized than other case representations due to the specific structure of problems and solutions in planning.By developing thefirst case-based planner(CHEF),Hammond(1989)proposed a case-based approach that uses a plan-like representation of cases.In planning,a problem is typically described by an initial state and a goal state.A solution is a totally or partially ordered sequence of actions.In principle,a planning case contains such a problem description and its related plan.The representation formalisms used in general AI planning significantly influence the case representations used.States are usually represented as sets of propositions in a logic language.The solution plan consists of a set of actions described,for instance,by operator terms,together with an ordering relation on them.A crucial step in case-based planning is the reuse phase,which requires strong plan adaptation capabilities.For plan adaptation,derivational analogy has demonstrated several advantages over transformational adaptation approaches.However,this method requires that additional information about successful and/or failed planning decisions be recorded as part of the case representation.Carbonell(1986)was thefirst to propose storing a complete derivational trace of the decision process in a planning case.Each derivational trace is augmented with an explicit justification structure explaining each individual decision.ReferencesAha,D,Kibler,D&Albert,M,1991.“Instance-based learning algorithms,”Machine Learning6:37–66. Bergmann,R,2002.Experience Management:Foundations,Development Methodology,and Internet-Based Applica-tions,Berlin:Springer.Bergmann,R&Wilke,W,1996.“On the role of abstraction in case-based reasoning,”in Smith,I&Faltings,B(eds.) Proceedings of the3rd European Workshop on Case-Based Reasoning,Berlin:Springer,pp.28–43. Bergmann,R&V ollrath,I,1999.“Generalized cases:Representation and steps towards efficient similarity assess-ment,”in Burgard,W,Christaller,T&Cremers,AB(eds.)KI-99:Advances in Artificial Intelligence,Berlin: Springer,pp.195–206.Carbonell,JG,1986.“Derivational analogy:A theory of constructive problem solving and expertise acquisition,”in Michalski,RS,Carbonell,JG&Mitchell,TM(eds.)Machine Learning:An Artificial Intelligence Approach (Volume II),Los Altos,CA:Morgan Kaufmann,pp.371–392.Gebhardt,F,V oß,A,Gr¨a ther,W&Schmidt-Belz,B,1997.Reasoning with Complex Cases,Kluwer Academic Publishers.G¨o ker,M&Roth-Berghofer,T,1999.“The development and utilization of the case-based help-desk support system HOMER,”Engineering Applications of Artificial Intelligence12(6):665–680.Hammond,KJ,1989.Case-Based Planning:Viewing Planning as a Memory Task,Boston:Academic Press. Hinrichs,TR,1992.Problem Solving in Open Worlds:A Case Study in Design,Northvale,New Jersey:Erlbaum. Kolodner,J,1983.“Maintaining organization in a long term dynamic memory,”Cognitive Science7:243–280. Kolodner,J,1991.“Improving human decision making through case-based decision aiding,”Artificial Intelligence Magazine12(2):52–68.Kolodner,J,1993.Case-Based Reasoning,San Mateo,California:Morgan Kaufmann.Leake,D&Wilson,DC,1999.“Combining CBR with interactive knowledge acquisition,”in Althoff,KD,Bergmann, R&Branting,LK(eds.)Proceedings of the3rd International Conference on Case-Based Reasoning,Berlin: Springer,pp.203–217.Lenz,M&Burkhard,HD,1996.“Case retrieval nets:Basic ideas and extensions,”in G¨o rz,G&H¨o lldobler,S(eds.) KI-96:Advances in Artificial Intelligence,Berlin:Springer,pp.227–239.Manago,M,Bergmann,R,Wess,S&Traph¨o ner,R,1994.CASUEL:A Common Case Representation Language, ESPRIT Project INRECA,No.6322,Deliverable D1,University of Kaiserslautern,Kaiserslautern Germany. Plaza,E,1995.“Cases as terms:A feature term approach to the structured representation of cases,”in Veloso,M& Aamodt,A(eds.)Proceedings of the1st International Conference on Case-Based Reasoning,Berlin:Springer, pp.265–276.Porter,BW,Bareiss,R&Holte,RC,1990.“Concept learning and heuristic classification in weak-theory domains,”Artificial Intelligence45:229–264.Zito-Wolf,R&Alterman,R,1992.“Multicases:A case-based representation for procedural knowledge,”in Proceedings of the14th Annual Conference of the Cognitive Science Society,Lawrence Erlbaum,pp.331–336.。