远程协同故障诊断关键技术及其应用研究

《远程协同故障诊断关键技术及其应用研究》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、西北工业大学博士学位论文远程协同故障诊断关键技术及其应用研究姓名：李盘靖申请学位级别：博士专业：机械制造及其自动化指导教师：秦现生20060601摘要并给出了基于该模型的智能故障检测与诊断系统的建立方法。最后，以FMS为主要算例，对上述模型及其相关算法进行了应用研究，相应的验证结果表明了理论研究成果的J下确性与可行性。在此基础上，设计了原型系统UN．E．CFD的开发方案，并给出了总结与展望。关键词：复杂装备、e．维护、远程协同故障诊断、博弈融合、行为智能本文部分研究内容受：欧盟AsiaLilll(国际合作项目“P

2、romotingandAssistingEngineeringDesigIlEducationinChiIla”(AsI，B7—301／98／679-09)博士生联合培养计划资助。一Il—两北f：业人学博十学位论文ResearchonKeyIssuesandApplicationofRemoteCooperativeFaultDiagnosis(PhDDissertation)NorthwestemPolytechnicalUniversity，Xi’aIl，P．R，ChinaABSTRACTNow跚

3、a，rS，

4、thecompJexequipmen￡sarepropeIledt0jarge-scaJe，fuIJautomat帆higllmnmngsPeed，whichleadsthefnltdete“ona11ddiagnosis(FDD)intoadilemmaundertraditionalmodeforthehjghcost，lowefnciencyandetc．Ther；cfore，remote，cooperative，intelligentmetllodsareintmducedimoFDDdomaina11d

5、becomet量lefesearchfocllsrecemly．T叮toaddresst上leissues捌senintheFDDfortIlecomplexsystem毗re-maintenanceenviromnent，misdisse删ion诅ket11eflexiblemallu触uringsystem(FMS)髂t量le印plicationobjec‘proposesanaItemativeschemeonmefemotecoope阳tiVeFDD，whichincludesfhmeworkcons扫m

6、ction，supponingtechniques，systemreaIizationand印pIicationstudy．First，b豁edontIlcanalyzingofrcsearchback伊oundofremoteFDD，melitefaturercviewa11dfheence巾“sessur坩y’tllisdisseftationpresentsproblemdescriptionafldputfonvardafemotecooperativeFDDfr锄eworkforthecomplexsy

7、stemundere-maintenanceenviro岫ent，whichistitJedwithE—CFD(e—CooperativeFDD)．Trytoaddresstheacquirement，transmission，analysisanddecisionofmulti-sourcesdiagnosticinfonnation，acenter-distributiOnFDDfhmeworkisconst兀lctedb船edontlleprocessimeFationofdevicet0e．busines

8、s(D2E)．Bya11alyzingthetheoreticprinciplea【ldkeytechnologjes，givesthelogicals仇JctIlrcaIld劬ctionalmodelofE—CFD，whichproVidesawaytotlleef话ctiVejntegrationofFDDsystemande．maintenanceenvironment．SecOnd，forthet船kdecompositionaJldcontrOlwithinrcmotecooperativeFDD，嘶s

9、djssenatjongjvesahybfjda】gorifhfnbasedonreIatjonmodeJandBayesnet们rk．111erelationmodelingmethodisintroducedtogives01utionofFDDpathpro鲈mlmingincoarseleveI．BasedonincidencematrixandBayesnetw