資源描述:
《基于小波包分解的滾動軸承故障信號頻域特征提取方法研究.pdf》由會員上傳分享,免費在線閱讀,更多相關(guān)內(nèi)容在應用文檔-天天文庫。
1、基于小波包分解的滾動軸承故障信號頻域特征提取方法研究丘世因,袁銳波(昆明理工大學機電工程學院,云南昆明650500)ResearchoftheRollingElementBearingFaultSignalFrequencyDomainFeatureExtractionMethodBasedontheWaveletPacketDecompositionQIUShi—yin,YUANRui—bo(FacultyofMechanicalandElectricalEngineering,KunmingUniversityofScie
2、nceandTechnology,Kunming650500,China)摘要:對軸承故障信號進行3層小波包分解,重cientsbywhichthebearingfaultsignalswerere—構(gòu)第3層所有節(jié)點,提取重構(gòu)信號頻譜的峰值作為constructed.Thepeakvaluesextractedfromthe故障特征點并構(gòu)成特征空間,計算特征空間的平均reconstructingsignalspectrumconstructedafea—歐氏距離,平均歐氏距離最小時對應的節(jié)點即為最turespace.Then,
3、theminimumaverageEuclidean優(yōu)小波包節(jié)點,重構(gòu)最優(yōu)節(jié)點得到最優(yōu)重構(gòu)信號并distancecalculatedfromthefeaturespaceindicated從中提取特征點構(gòu)成最優(yōu)特征空間,最后,對最優(yōu)特theoptimalwaveletpacketnode.Theoptimalfea—征空間進行K均值聚類。對4種轉(zhuǎn)速下軸承的4turespacecouldbeconstructedbythefeature種狀態(tài)進行特征提取與模式識別試驗,結(jié)果表明,運pointsextractedfromthesi
4、gnalsreconstructedby用該方法能有效提取軸承故障的特征,并使故障特theoptimalwaveletpacketnodes.Finally,theopti—征空間具有最低的類內(nèi)離散度,獲得了較高的模式malfeaturespacewasusedfortheK—meansclus—識別準確率。tering.Thefeatureextractionandpatternrecogni—關(guān)鍵詞:軸承故障;特征提取;小波包分解;最優(yōu)tiontestofthefourkindsofbearingconditionsun—
5、節(jié)點;K均值聚類;模式識別derfourkindsofrotationspeedswasdetailed.The中圖分類號:TH133;TP391testresultsshowthismethod,whichcanextract文獻標識碼:Athebearingfaultfeatureefficientlyandmakethe文章編號:1001—2257(2014)05—0012—05faultfeaturespacehavethe1owestwithin—classAbstract:Firstofall,thebearing
6、faultsignalsscatter,wonsahighpatternrecognitionaccuracy.weredecomposedintothreelayerswaveletcoeffi一Keywords:bearingfault;featureextraction;waveletpacketdecomposition;optimalnode;K—收稿日期:2013—11—06基金項目:云南省自然科學基金重點項目(2010CD030)meansclustering;patternrecognitiontirecath
7、eterinsertion:algorithmsandexperiments[J].TheInternationalJournalofRoboticsResearch,[J].TheInternationalJournalofRoboticsResearch,2010,29(13):1661—1683.2009,28(9):I101—1117.[9]SearsP,DupontP.AsteerableneedletechnologyusingE7]CaminoWW.Activetipcontrolforsmalldiameter
8、curvedconcentrictubes[A].IEEE/RSJInternationalcatheters:Powertransmissionandtransductionre~ConferenceonIntelligentRobotsandSystems[C].quir