毕设外文翻译.docx

毕设外文翻译.docx

《毕设外文翻译.docx》由会员分享，可在线阅读，更多相关《毕设外文翻译.docx（13页珍藏版）》请在冰豆网上搜索。

毕设外文翻译

毕设外文翻译

英文翻译

题目:

通过分析变压器中溶解气体而进行故障诊断的专家系统

姓名:

宋日成

学院:

工学院

专业:

自动化

班级:

自动化112班

学号:

指导教师:

陆静职称:

讲师

2015年3月14日

南京农业大学教务处制

AnExpertSystemforTransformerFaultDiagnosisUsingDissolvedGasAnalysis

W.S.Chan·Y.L.Xu·X.L.Ding·W.J.Dai

Received:

9November2005/Accepted:

11August2006/Publishedonline:

7September2006Springer-Verlag2006

AbstractInordertoautomatethetransformerfaultdiagnosis,improvetheaccuracyofjudgment,theintroductionofartificialintelligenceexpertsystemfaultdiagnosis.Thesystemisbasedonfuzzyreasoningconfidencebyobservingtheinformation,theuseofknowledgetoreachaconclusion.Testprovedthismethodreducestherandomnessofjudgment,improveddiagnosis.

Keywords:

ExpertSystemTransformerTroubleshooting

1Introduction

Thepowertransformerisamajorapparatusinapowersystem,anditscorrectfunctioningisvitaltosystemoperations.Inordertominimizesystemoutages,manydeviceshaveevolvedtomonitortheserviceabilityofpowertransformers.Thesedevices,suchas,Buchholzrelaysordifferentialrelays,respondonlytoaseverepowerfailurerequiringimmediateremovalofthetransformerfromservice,inwhichcase,outagesareinevitable.Thus,preventivetechniquesforearlydetectionfaultstoavoidoutageswouldbevaluable.Inthisway,analysisofthemixtureofthefaultygasesdissolvedininsulationoilofpowertransformerhasreceivedworldwiderecognitionasaneffectivemethodforthedetectionofincipientfaults.ManyresearchersandelectricalutilitieshavereportedontheirexperienceanddevelopedinterpretativecriteriaonthebasisofDGA.However,criteriatendtovaryfromutilitytoutility.Eachapproachhaslimitationsandnoneofthemhasafirmmathematicaldescription.Therefore,transformerdiagnosisisstillintheheuristicstage.Forthisreason,knowledge-basedprogrammingisasuitableapproachtoimplementinsuchadiagnosticproblem.BasedontheinterpretationofDGA,aprototypeofanexpertsystemfordiagnosisofsuspectedtransformerfaultsandtheirmaintenanceproceduresisproposed.Thesignificantsourceinthisknowledgebaseisthegasratiomethod.Somelimitationsofthisapproachareovercomebyincorporatingthediagnosticprocedureandthesyntheticexpertisemethod.Furthermore,databasesadoptedfromTPCSgasrecordsoftransformersareincorporatedintotheexpertsystemtoincreasethepracticalperformance.Uncertaintyofdiagnosisismanagedbyusingfuzzysetconcepts.Thisexpertsystemisconstructedwithrulebasedknowledgerepresentation,sinceitcanbeexpressedbyexperts.Theexpertsystembuildingtool,KnowledgeEngineeringSystem（KES）,isusedinthedevelopmentoftheknowledgesystembecause,ithasexcellentman-machineinterfacethatprovidessuggestions.Moreover,itsinferencestrategyissimilartotheMYCIN.Afamousrule-basedexpertsystemusedformedicaldiagnosis.Theuncertaintyofhumanqualitativediagnosticexpertise,.,keygasanalysis,andanotherquantitativeimprecision,suchas,normsthresholdandgasratioboundariesetc.,aresmoothedbyappropriatefuzzymodels.Withtheresultsofsuchimplementation,differentcertaintyfactorswillbeassignedtothecorrespondingexpertisevariables.Bothevent-driven（forwardchaining）andgoal-driven（backwardchaining）inferencesareusedintheinferenceenginetoimprovetheinferenceefficiency.Todemonstratethefeasibilityoftheproposedexpertsystem,aroundhundredsofTPChistoricalgasrecordshavebeentested.Itisfoundthatmoreappropriatefaultytypesandmaintenancesuggestionscansupportthemaintenancepersonalstoincreasetheperformanceoftransformerdiagnosis.

OFDIAGNOSISANDINTERPRETATION

Likemanydiagnosticproblems,diagnosisofanoil-immersedpowertransformerisaskilledtask.Atransformermayfunctionwellexternallywithmonitors,whilesomeincipientdeteriorationmayoccurinternallytocauseafatalprobleminthelatterdevelopment.AccordingtoaJapaneseexperience,nearly80%ofallfaultsresultfromincipientdeteriorations.Therefore,faultsshouldbeidentifiedandavoidedattheearliestpossiblestagebysomepredictivemaintenancetechnique.DGAisoneofthemostpopulartechniquesforthisproblem.Faultgasesintransformersaregenerallyproducedbyoildegradationandotherinsulatingmaterials,.,celluloseandpaper.Theoretically,ifanincipientoractivefaultispresent,theindividualdissolvedgasconcentration,gassingrate,totalcombustiblegas（TCG）andcellulosedegradationareallsignificantlyincreased.Byusinggaschromatographytoanalyzethegasdissolvedinatransformer'sinsulatingoil,itbecomesfeasibletojudgetheincipientfaulttypes.Thisstudyisconcernedwiththefollowingrepresentativecombustiblegases;hydrogen（H2）,methane（C2H2）,ethane（C2H6）,ethylene（C2H4）,acetylene（C2H2）andcarbonmonoxide（C0）.ManyinterpretativemethodsbasedonDGAtodiagnosethenatureofincipientdeteriorationhavebeenreported.Evenundernormaltransformeroperationalconditions,someofthesegasesmaybeformedinside.Thus,itisnecessarytobuildconcentrationnormsfromasufficientlylargesamplingtoassessthestatistics.TPCinvestigatedgasdatafrompowertransformerstoconstructitscriteria.Thedevelopedknowledgebaseinthispaperispartiallybasedonthesedata.Ontheotherhand,Dornerburgdevelopedamethodtojudgedifferentfaultsbyratingpairsofconcentrationsofgases,.,CH/H,CH/C2H4,withapproximatelyequalsolubilityandfusioncoefficients.Rogersestablishedmarecomprehensiveratiocodestointerpretthethermalfaulttypeswiththeoreticalthermodynamicassessments.Thisgasratiomethodwaspromisingbecauseiteliminatedtheeffectofoilvolumeandsimplifiedthechoiceofunits.Moreover,itsystematicallyclassifiedthediagnosisexpertiseinatableform.Table1displaystheratiomethodasproposedbyRogers.Thedissolvedgasmayvarywiththenatureandseverityofdifferentfaults.Byanalyzingtheenergydensityoffaults,it'spossibletodistinguishthreebasicfaultprocesses:

overheating（pyrolysis）,corona（partialdischarge）andarcingdischarge.Coronaandarcingarisefromelectricalfaults,whileoverheatingisathermalfault.Bothtypesoffaultsmyleadtodeterioration,whiledamagefromoverheatingistypicallylessthanthatfromelectricalstress.Infect,differentgastrendsleadtodifferentfaultytypes,thekeygasmethodisidentified.'ForexamplelargeamountsofCHandHareproducedwithminorarcingfault4.quantitiesofCH2aidC2H42maybeasymptomofanarcingfault.

PROPOSEDDIAGNOSTICEXPERTSYSTEM

Thisstudyisaimedatdevelopingarule-basedexpertsystemtoperformtransformerdiagnosissimilartoahumanexpert.Thedetailsofsystemprocessingaredescribedbelow.

TheProposedDiagnosticMethod

Diagnosisisataskthatrequiresexperience.Itisunwisetodetermineanapproachfromonlyafewinvestigations.Therefore,thisstudyusesthesyntheticexpertisemethodwiththeexperiencedproceduretoassistthepopulargasratiomethodandcompletepracticalperformance.

ExperiencedDiagnosticProcedure

Theoverallprocedureofroutinemaintenancefortransformersislisted.ThecoreofthisprocedureisbasedontheimplementationoftheDGAtechnique.Thegasratiomethodisthesignificantknowledgesource.Someoperationallimitationsofthegasratiomethodexist.Theratiotableisunabletocoverallpossiblecases.Minimumlevelsofgasesmustbepresent.ThesolidinsulationinvolvingCOandCOarehandledseparatelyandthegasratiocodeshavebeendevelopedmainlyfromafree-breathingtransformer.Otherdiagnosticexpertiseshouldbeusedtoassistthismethod.Norms,syntheticexpertisemethodanddatabaserecordshavebeenincorporatedtocompletetheselimitations.ThefirststepofthisdiagnosticprocedurebeginsbyaskingDGAforanoilsampletobetested.Moreimportantrelevantinformationaboutthetransformer'scondition,suchasthevoltagelevel,thepreservativetype,theon-line-tap-changer（OLTC）state,theoperatingperiodanddegassedtimemustbeknownforfurtherinference.Noms（criteria）SetupbyTPCpowertransformers'gascharacteristicdataarethenusedtojudgethetransformers'cactivefaultispresent,theindividualdissolvedgasconcentration,gassingrate,totalcombustiblegas（TCG）andcellulosedegradationareallsignificantlyincreased.Byusinggaschromatographytoanalyzethegasdissolvedinatransformer'sinsulatingoil,itbecomesfeasibletojudgetheincipientfaulttypes.Thisstudyisconcernedwiththefollowingrepresentativecombustiblegases;hydrogen（H2）,methane（C2H2）,ethane（C2H6）,ethylene（C2H4）,acetylene（C2H2）andcarbonmonoxide（C0）.ManyinterpretativemethodsbasedonDGAtodiagnosethenatureofincipientdeteriorationhavebeenreported.Evenundernormaltransformeroperationalconditions,someofthesegasesmaybeformedinside.Thus,itisnecessarytobuildconcentrationnormsfromasufficientlylargesamplingtoassessthestatistics.TPCinvestigatedgasdatafrompowertransformerstoconstructitscriteria.Thedevelopedknowledgebaseinthispaperispartiallybasedonthesedata.Ontheotherhanondition.Fortheabnormalcases,thegasratiomethodisusedtodiagnosetransformerfaulttype.Ifdifferentorunknowndiagnosisresultsarefoundfromtheseratiomethods,afurthersyntheticexpertisemethodisadopted.Aftertheseprocedures,differentseveritydegreesareassignedtoallowappropriatecorrespondingmaintenancesuggestions.

SyntheticExpertiseMethod

Theratiotrend,normst