資源描述：
《Automatic Classification of Power Quality》由會(huì)員上傳分享，免費(fèi)在線閱讀，更多相關(guān)內(nèi)容在學(xué)術(shù)論文-天天文庫(kù)。
1、IEEETRANSACTIONSONINDUSTRIALELECTRONICS,VOL.61,NO.1,JANUARY2014521AutomaticClassi?cationofPowerQualityEventsUsingBalancedNeuralTreeB.Biswal,Member,IEEE,M.Biswal,Member,IEEE,S.Mishra,SeniorMember,IEEE,andR.JalajaAbstractThispaperproposesanempirical-modedecomposi-time.TheFouriertransf
2、ormgivesinformationregardingthetion(EMD)andHilberttransform(HT)-basedmethodforthefrequencycomponentspresentbutdoesnotcontaininformationclassi?cationofpowerquality(PQ)events.Nonstationarypoweronwhentheyexistandforhowlong.AlthoughtheFouriersignaldisturbancewaveformsareconsideredasthes
3、uperimposi-transformisoneofthefasttechniques,itsef?ciencyislimitedtionofvariousundulatingmodes,andEMDisusedtoseparateouttheseintrinsicmodesknownasintrinsicmodefunctions(IMFs).tostationarysignalsonly.MostPQeventsarenonstationaryandTheHTisappliedonalltheIMFstoextractinstantaneousam-he
4、ncerequiretechniquethatwouldnotonlyprovidefrequencyplitudeandfrequencycomponents.Thistimefrequencyanalysisinformationbutalsocapturethetimingofoccurrenceoftheresultsintheclearvisualdetection,localization,andclassi?cationdisturbance.Theshort-timeFouriertransformgivesagoodofthedifferen
5、tpowersignaldisturbances.Therequiredfeaturecharacterizationofthesignal.Itprovidesfrequencyaswellasvectorsareextractedfromthetimefrequencydistributiontoperformtheclassi?cation.Abalancedneuraltreeisconstructedtotimeinformation.Thenonstationarynatureofthesignaliswellclassifythepowersig
6、nalpatterns.Finally,theproposedmethodisde?ned.However,duetotheconstantwindowlength,somecomparedwithanS-transform-basedclassi?ertoshowtheef?cacycharacteristicsofthesignalarenotdetectedwell.ThetimeoftheproposedtechniqueinclassifyingthePQdisturbances.andfrequencyresolutionislimitedbyth
7、eHeisenberg–GaborIndexTermsBalancedneuraltree(NT)(BNT),empirical-inequality.Differenttypesofdisturbanceswouldrequirewin-modedecomposition(EMD),Hilberttransform(HT),dowsofdifferentlengths.Choosingthebestwindowlengthinstantaneousfrequency(IF),intrinsicmodefunction(IMF),couldbeaproblem
8、.Thewavelettransfor
