new procedure for gear fault detection and diagnosis using instantaneous angular speed 2017 bing li参考.pdf

该【new procedure for gear fault detection and diagnosis using instantaneous angular speed 2017 bing li参考 】是由【小舍儿】上传分享，文档一共【14】页，该文档可以免费在线阅读，需要了解更多关于【new procedure for gear fault detection and diagnosis using instantaneous angular speed 2017 bing li参考 】的内容，可以使用淘豆网的站内搜索功能，选择自己适合的文档，以下文字是截取该文章内的部分文字，如需要获得完整电子版，请下载此文档到您的设备，方便您编辑和打印。:..MechanicalSystemsandSignalProcessing85(2017)415–428ContentslistsavailableatScienceDirectMechanicalSystemsandSignalProcessingjournalhomepage:ate/ymsspearfaultdetectionanddiagnosisusinginstantaneousangularspeedBingLi,XiningZhangn,JiliWuStateKeyLaboratoryforManufacturingSystemEngineering,Xi’anJiaotongUniversity,710049Xi’an,PRChinaarticleinfoabstractArticlehistory:plexityofgeardynamics,thefaultdiagnosisresultsintermsofReceived14March2016vibrationsignalaresometimeseasilymisledandevendistortedbytheinterferenceofponentslikebearings,,theresearchfield27June2016ofInstantaneousAngularSpeed(IAS)'sadvantages,biningtheEmpiricalModeKeywords:position(EMD)andAutocorrelationLocalCepstrum(ALC),asapre-processingstep,(IMFs).Nevertheless,,,thesensitiveIMFisstillin-,asthefinalstep,ALCisusedforthepurposeofsignalde-.,thehighlightedquefrencyanditsrahmonicscor-respondingtotherotaryperiodanditsmultiplearedisplayedclearlyinthecepstrumrecordoftheproposedmethod.&-,inpractice,vibration-,vibrationsignalsacquiredfromamultistageelerometerisalwaysmountedontheoutersurfaceofthegearboxcasenearthemainbearinghousing[1,2].Duetothefactorsmentionedabove,usefulinformationisinevitablycorrupted,whichmakesitdifficulttodiagnosegearboxesfromsuchvibrationsignals[3].Thus,://dx./.-3270/&.:.../MechanicalSystemsandSignalProcessing85(2017)415–,theresearchfieldofInstantaneousAngularSpeed(IAS)hasattractedsignificantattentionfromtheresearchcommunity,whichisattributabletoitsownadvantagesoverconventionalvibrationanalysis[1,4,5].[6].Arisingfromthesebenefits,differentadvancedIAS-basedtechniqueshavebeendevelopedasionssuchasin:rotors[4],dieselengines[7,8],bearings[9],inductionmotors[4,6],plexityandnon-stationary,,duetothehighsamplefrequencyincollectingtherawsignal,agreatdealofdatacanbegenerated,,itisnoticeablethatthepeaksaboveacertainthresholdaredramaticallyreduced,whichmeanstheIASsignalisover-,manypragmaticidentificationapproachessuchastime–frequencyrepresentation,empiricalmodede-composition(EMD),wavelettransform,essfullyestablishedtoenhancethesignalfeatureandsimultaneouslyprovidepotentproofsformaintenancedecision-making[10–14].Amongthesediagnostictechnologies,position(EEMD)havebeenwidelyusedtoanalyzevibrationsignalsinthefieldoffaultdiagnosis,particularlyforsignalsinnon-stationarycasesandsignalswithalargersize[11,15].EEMD,developedbyWuandHuang[16],isabletoeliminatetheproblemofmodemixingeffectivelybymeansofaddingfinitewhitenoisetotherawsignal,,poseasignalintodifferentintrinsicmodefunctions(IMFs),,-lectionofIMFsareusuallyreliedontheuser',ietal.[11],[15],,,forthismethod,,recentlytwoimprovedcepstrumshavebeendevelopedessively[17,18].BothLocalCepstrum(LC)analysisanditsimprovedformAutocorrelationLocalCepstrum(ALC)focusonachievingthemaximumutilityofthelocal(partial),-ponents[18].However,al-,thenoisereductionperformanceofLCisalsolimitedsinceitdoesnotconcerntoomuchaboutthesignalde-,biningLCwithau-essfullyutilizedtodealwiththegearboxdiagnosisusingvibrationsignals[17].OntheconsiderationofIASspectrumsignatures,,inthispaper,,signalreconstructionisdeliveredasapre-,,thecosinesimilaritymeasure[19,20]thatismostwidelyreportedforsimilaritymea-,somerelevanttheoriesthatsupporttheproposedmethod,includingsignalreconstruction,EMD,ALC,,thefusionmethodisappliedtothesetwogearboxesunderdifferenthealthstatusesandsomeresultsanddis-,,,,binationof:.../MechanicalSystemsandSignalProcessing85(2017)415–,electricsupply,filteringanddigitalization[4,21].Besides,agearwithdefectswilltheoreticallygenerateanunevendistributionoftangentialforcebetweenthemeshingteeth,,aspeedfluctuationhappens,,theIASsignalisactuallynon-stationarywiththecharacteristicsoffrequencymodulation(FM),signalsderivedfromangulardomainwithstronganti-disturbanceabilityandgreatrobustness,ponentslikebearings,,theIASsignalhasthehighnoisesuppressioncapabilityandgreatperformancesincethisalgorithmistheoreticallydevelopedonthebasisoffrequencymodulation,,[22],measurementperformanceofIASsignalsisbasedontwobasicprinciples::2π60fhω=?NfNi()1wherefhrepresentstheclockfrequencyorsamplefrequency;Niisthenumberofdatapointsbetweentworisingedgesoftheencodersignal;'[4],=×++4????fhf()????smin00()2wherefsmin,f0,PandNhrespectivelydenotetheminimalsamplefrequency,therotationfrequency,-,theefficiencyofIAS-basedsignalessentiallyreliesonseveralfactorslikeproperestimationalgorithm,sensorresolutionandirregularity,interpolationalgorithms,frequencyaliasing,,-processingDuetothehighresolutionoftheangularsensor,.,opticalencoder,,ahighsamplefrequencyisalwaysneededandhencealargeamountofdataaregeneratedin-evitably,,byspectralobserving,wecanclearlyfindtheoversampledphenomenoninthespectrumwhereaboveacertainthreshold,-,asthepre-processingmethod,aneffectivestrategyofsignalre-,thefrequencyresolutionremainsthesame,pletelypreservedforthefurtherinvestigations.:.../MechanicalSystemsandSignalProcessing85(2017)415–,EMDmethodiswell-developedandwidelyappliedtotheanalysisofnon-linearandnon-,,eachIMFmustsatisfythefollowingconditions[15].1)Inalldataset,thenumberofextremesandthenumberofzero-)Atanypoint,[15].Anyrawsignalxt()canposedbytheEMDalgorithmasfollows:1)Afterextractingallthelocalextremes(includingmaximaandminima)ofxt(),)hemeansignalasmt1(),thedifferencebetweenxt()andmt1()ponenth1(t),.,h11()()()txtmt=?.3)Replacext()byh1(t),figureoutwhetherh1(t)meetstheIMFconditionsandifnot,repeatsteps(1)and(2)untilh1(t),ctht11()=())Separatect1()fromxt()togettheresidualsignalrt1())Derivationwillcontinueuntilthestoppagecriterionofthesignal'positionisfulfil