
中文摘要中文摘要随着电力电子技术的发展,电力系统中整流器、变频调速装置、电弧炉及各种电力电子装置设备不断增加,这些非线性、冲击性和不平衡负荷的用电特性,对供电质量造成了严重污染,给系统的安全、稳定、经济运行造成了严重影危害;另一方面现代化工业、商业及居民用户的用电设备对电能质量更加敏感,对电能的使用和需求提出了更高的要求。如何才能把谐波的危害最大限度的降低,是目前电力系统领域极为关注的问题,解决这一问题的关键在于准确掌握谐波的次数、幅值等等。无谐波已经成为“绿色"电能的主要标志,谐波含量己成为衡量电能质量的一个重要参数,精确掌握谐波含量对于抑制电网谐波是具有重要的意义。电力负荷是电力系统的重要组成部分,负荷建模一直是电力界公认的一个难题,受到学者的广泛关注。精确负荷谐波模型对电力系统谐波分析和计算有重要的意义,从而进一步影响到电力系统的监控、管理和治理。与潮流计算相比,谐波分析对元件的模型的精度要求更高。与发电机、变压器和输电线路的谐波建模相比,负荷谐波建模进展缓慢,粗糙的负荷谐波模型阻碍了整个电力系统仿真精度。因此,精确负荷谐波模型的建立具有重要的意义。本文首先系统而简要的介绍电力系统谐波的基本概念、危害、来源及几种谐波分析方法应用现状及存在的问题,针对以往算法在应用中存在的计算精度低、计算量大等问题,提出建立基于自适应神经模糊推理系统谐波分析模型,研究了基于BP梯度下降法和最小二乘法结合的算法,并通过与其他算法仿真结果的对比验证了该算法的有效性,仿真结果表明,本文提出的电力系统谐波分析方法具有计算精度高和计算速度快的特点。
其次,本文叙述了电力负荷建模意义、现状及支持向量机算法原理,针对通用的负荷谐波模型,提出了一种新的基于最4,-乘支持向量机的参数辨识方法,并通过与其他参数辨识结果的对比,验证了算法的精度,仿真结果表明本文提出的负荷谐波模型参数辨识算法是可行且有效的。针对不对称三相系统的负荷谐波建模问题,研究了适用于各种三相负荷的谐波建模模型的对称分量算法,并通过算例验证了算法的可行性。最后,对本文的研究内容作了总结,并展望谐波分析和负荷谐波建模的研究方向。关键词:电力系统;谐波分析;负荷建模;参数辨识;自适应神经模糊推理系统;最小二乘支持向量机;对称分量法;分类号:TM711ABSTRACTABSTRACTof electromics moreWiththe and thepower technique,moredevice,developmentsuchas arcfumaceandSOonare intoconverter,electricrectifier,frequency appliedsortofloadisthis powerqualitypowersystem,asbecomsworsethan the of andbefore,whichendangerssafety,stabilitypowersystemthe of theother deviceusedinmoderndecreaseseconomyoperation.Onhand,theand life,is influencedindustry,commencepeople’Sdaily easily bypowerquality,issueof isconcernedallofUS.Howtotherefore,theimprovingpowerquality byisanissue thefieldminimizethe oftheharmonics of congerllto ofdamage great powertosolutionofthisissueliesinthe determinationofthesystems.Thekey quantitativeof hasbecomeand harmoniouswaves.Nonharmoniccomposition,magnitudephasesthemain of contentofharmonicin isthesymbolgreen,the powersystem importantin harmonic is insignificancepowerparameterpowerquality,SO ysisimportantsystem.isan constituentoftheelectrical itsThe loadpower important powersystem,andworkhas adifficult ofthe themodeling alwaysbeing problempowersystem,acquiringdomesticand attation.Accuratemodelisforeignscholars’widespreadpower significantinthe harmonic and furtheraffectsthepowersystem ysiscalculation,therebyand haspowersystemmonitoring,managementgovemance.Harmonicysishigheronthe ofthe thanthe wimmodel flow.Comparedprecision components powerandtransmissionline ofthe slowgenerators,transformers modelingharmonic,theinharmonic ofelectric the harmonic ofprogress modeling loads,andrough modelingelectricloadshinderedthe oftheentireaccuracy powersystemthe harmonic ofelectricloadsisofprecise modeling greatsignificance.the of ofandbriefintroduces harmonic,hazardsystemic conceptFirstly,thepaperofthetraditionalandnewharmonicharmonic,harmonicsource,existedproblemare several ofwhichused theviewofthealgorithms problemsysis frequently.Indifferent suchasthelow andthealgorithms accuracy bigcomputationcomputationartificial basedandSO networkbasedon Neural—Networkquantityon,all adaptive FuzzyInferencewas inthe withtheharmonicmodeloftheelectricSystempresentedpaperthe basedontheBP descentruleandleastpowersystem,andalgorithm gradient squareswereresearched.To inthe ofmethod ofthe fieldvalidity algorithm powersystemABSTRACTharmonic simulationandotherysis,the researchcomputer algorithmscomparisonwere inthe researchresultsshowedthatgiven paper.The theharmonicysisoftheapproach inthe hadthepowersystem andthepaper highcomputationaccuracyfastcomputationspeed.ofload status andSecondly,thesignificance vectormodeling,thequo supportmachine are theviewalgorithm ofthe load newreviewed,in generalizedmodel,aidentificationbasedonleast vectorparameter algorithm squares machinewassupportinthe ofthepresented simulationandotherpaper,tovalidityalgorithm,thecomputerresearchalgorithmscomparisonwere inthe theresearchresultsgiven paper.Andshowedthatthe identificationinparameter the isfeasibleandeffective.algorithmpaperthe ofviewtheharmonic ofelectricofThirdly,in loadsmodeling asymmetricload,the issuitableforall ofproposedalgorithm typesmodelingthree-phaseload,thesimulationresultsindicatethatthe has andproposedalgorithmgreatvalidityaccuracy.summarizetheFinally,the makesa onPaper studycontent,moreover,itprospectwhich theharmonicandharmonic ofelectricinapplying ysis modeling loadsthepowersystem.KEYWORDS:PowerSystem;Harmonicysis;loadmodel;Parameteridentification;Neural—Networkbased InferenceANIFS Adaptive Fuzzy System ;LS.SVM Leastvectorsquaressupportmachines ;Symrnetricalcomponent;CLASSNO:TM711:独创性卢明独创性声明本人声明所呈交的学位论文是本人在导师指导下进行的研究工作和取得的研究成果,除了文中特别加以标注和致谢之处外,论文中不包含其他人已经发表或撰写过的研究成果,也不包含为获得北京交通大学或其他教育机构的学位或证书而使用过的材料。
本文来自电脑杂谈,转载请注明本文网址:
http://www.pc-fly.com/a/tongxinshuyu/article-29327-1.html
我的6是9