欢迎来到天天文库
浏览记录
ID:36809157
大小:2.30 MB
页数:58页
时间:2019-05-15
《基于小波包和支持向量机的机车轴承故障诊断研究》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、基于粒子群最小二乘支持向量机的故障诊断算法研究AbstractRailwaylocomotivebearingisakeycomponentoftransportationequipment,Asalongtime,withthehigh-loadoperation,thelocomotivebearingispronetoinjury.LotsofRailwaylocomotivebearingfaultsbelongstotheexistenceoroccurrenceoffailures,thence,thequalityofbearingrunni
2、nglocomotivesuperiorperformanceoftheentireplay.In.viewofthelocomotivebearingthespecialstatus,therefore,cartingoutOntherailwaylocomotivebearingfaultdiagnosishavegoodpracticalsignificance.硒ethesiswasmainlybasedOilwaveletpacketanalysisandsignalpcocessingsupportvectormachinebasedintel
3、ligentfaultdiagnosisresearchintwoareas.Themaincontentsareasfollows:Discussionandanalysisofthelocomotivebearingthevibrationcharacteristicsofvibrationsignalsandthemechanismof.failure,simulatedbearingfaultsignalsofdifferentcomponentsonthebasisofmasteringthefaultsignalcharacteristics,
4、whichcanprovidethemethodandbasisvectorsoflocomotivebearingfaultdiagnosis.Traditionalsignalprocessingmethodisonlysuitableforstationary,non-time-varyingsignalprocessing,anddoesnothavethecapabil/tyoflocalsignalanalysis,usingtheadvantagesofwaveletpacketanalysisinmutationdetectionandsi
5、gnaldenoisingtOdecomposewaveletpacketforfaultsignal,SOthatthecharacteristicsoffaultsignalsofdifferentfrequencyinformationcanbediscovered.Inordertoanalyzetheamountofbearingonwhetherthepo硫ofviewthavefaultornotwhiletakingthecomplexityofcomputationintobandsoffaultcharacteristicsaccoun
6、t,selectingobviously,whichcombined稍ththeideaoffrequencybandenergyanalysistoachievethefeaturevectorextractionofsimulatedfaultsignal’Sspecificband011bearing’outerring,innerringandtherolling.Usingleastsquare,ssupportvectormachinesforfaultdiagnosistrainingsamplesdonotneedalotoftrouble
7、,whichalsohasaclassificationabm【yinthecaseofhighdimension,goodhowever,thesupportvectormachineparameters(penaltyfactorandtheradialbasisfunction)modelassociationhaveagreaterimpactOilthediagnosticclassificationrate,witllthedeepstudyofparticleswal'malgorithmandalargenumberofsimulation
8、experiments,inordertoimprovethecl
此文档下载收益归作者所有