資源描述:
《Fault Detection by Means of Hilbert–Huang Transform of the Stator Current in a PMSM With Demagnetization 》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在教育資源-天天文庫(kù)。
1、312IEEETRANSACTIONSONENERGYCONVERSION,VOL.25,NO.2,JUNE2010FaultDetectionbyMeansofHilbertHuangTransformoftheStatorCurrentinaPMSMWithDemagnetizationAntonioGarciaEspinosa,JavierA.Rosero,JordiCusido,LuisRomeral,andJuanAntonioOrtega′AbstractThispaperpresentsanovelmethodtodiagn
2、osede-thecouplingbetweenthenonlinearmagneticandelectriccir-magnetizationinpermanent-magnetsynchronousmotor(PMSM).cuits,istakenintoaccountforthemotoranalysistopredicttheSimulationshavebeenperformedby2-D?nite-elementanalysisinperformancecharacteristicsofasystem,especiallywh
3、entheordertodeterminethecurrentspectrumandthemagnetic?uxsystemisunderfaultconditions[11].distributionduetothisfailure.Thediagnosticjustbasedonmotorcurrentsignatureanalysiscanbeconfusedbyeccentricityfail-ThefastFouriertransform(FFT)ofthestatorcurrenthasurebecausetheharmoni
4、ccontentisthesame.Moreover,itcanbeenappliedtodetectdemagnetizationfaultsbyanalyzingspe-onlybeappliedunderstationaryconditions.Inordertoover-ci?charmonics.However,itcannotbeappliedtononstationarycomethesedrawbacks,anovelmethodisusedbaseduponthesignals.Moreover,FFTanalysisc
5、annotdifferentiateharmonicsHilbertHuangtransform.Itrepresentstime-dependentseriesinaduetodemagnetizationfromothersduetoeccentricity[1],[12].2-Dtimefrequencydomainbyextractinginstantaneousfrequencycomponentsthroughanempirical-modedecompositionprocess.Timefrequencyanalysism
6、ethodshavebeenusedfornon-Thistoolisappliedbyrunningthemotorundernonstationarystationarysignalfeatureextraction,althoughsuccessfulappli-conditionsofvelocity.Theexperimentalresultsshowthereliabilitycationofthesetechniquesrequiresunderstandingoftheirre-andfeasibilityofthemet
7、hodologyinordertodiagnosethedemag-spectivelimitations.TheselectionofasuitablewindowsizeisnetizationofaPMSM.requiredwhenapplyingtheshort-timeFouriertransform(STFT)IndexTermsDemagnetization,empirical-modedecompositiontomatchwiththespeci?cfrequencycontentofthesignal,which(EM
8、D),faultdiagnosis,?nite-elementanalysis(FEA),Hilbertisgenerallynotknownapriori.Moreover,thereisa