摘要
Thispaperpresentsanovelapproachtodetectanddiagnosefaultsinthedynanmicpartofachisofstochasticsys-tems.theSuchagroupofsystemsaresubjectedtoasetofcrispinputsbuttheoutputsconsideredarethemeasurableprobabilitydensityfunctions(PDFs)ofthesystemoutput,ratherthanthiesystemoutputalone.Anewapproximationmodelisdevelopedfortheoutputprobabilitydensityfunctionssothatthedynamicpartofthesystemisdecoupledfrontheoutputprobabilitydensityfunctions.Anonlinearadaptiveobserverisconstructedtodetectanddiagnosethefaultinthedynamicpartofthesystem.Convergencyanalysisisperfomedfortheerrordynamicsraisedfromthefaultdetectionanddiagnosisphaseandanapplicabilitystudyonthedetectionanddiagnosisoftheunexpectedchangesinthe2Dgrmmagedistributionsinapaperformingprocessisincluded.
出版日期
2003年01月11日（中国期刊网平台首次上网日期，不代表论文的发表时间）
