基于相空间重构与平稳子空间分析的滚动轴承故障诊断.pdfVIP

振动与冲击

第34卷第22期JOURNALOFVIBRATIONANDSHOCK

基于相空问重构与平稳子空间分析的滚动轴承故障诊断

刘尚坤，唐贵基，庞彬

(华北电力大学机械工程系，河北保定071003)

摘要：针对如何分离淹没在与转子转速相关的背景信号及其它噪声中的滚动轴承故障特征信号问题，提出将相

空间重构与平稳子空间分析相结合的滚动轴承故障诊断方法。用相空间重构实现对滚动轴承故障振动信号升维；用平稳

子空间对高维信号中平稳、非平稳源信号进行区分，并对峭度值最大的非平稳源信号进行最小熵解卷积降噪；对降噪信号

进行包络谱分析提取轴承故障特征频率。仿真信号、故障诊断实例表明，诊断效果优于基于EMD的包络解调方法。

关键词：相空间重构；平稳子空间分析；最小熵解卷积；滚动轴承；故障诊断

中图分类号：TH133．3；TH165+．3文献标志码：ADOI：10．13465／j．cnki．jvs．2015．22．033

Faultdiagnosisforrollingbearingsbasedonphasespace

reconstructionandstationarysubspaceanalysis

LIUShang—kun，TANGGui-ji，PANGBin

(MechanicalEngineeringDepartment，NorthChinaElectricPowerUniversity，Baoding071003，China)

Abstract：Rollingbearing’Sfaultfeaturesignalsarenon—st~ionary，transientandoftensubmergedinbackground

signalsassociatedwiththecomponentsofrotatingspeedandothernoises．Howtoseparatethefaultfeaturesignalsfromthe

rollingbearing’Sblindsourcesisanimportantissue．Amethodcombiningthephasespacereconstructiontechniquewith

thestationarysubspaceanalysis(SSA)wasproposed．Thefaultvibrationsignal’Sdimensionwasincreasedbythephase

spacetechniqueandthenon-stationarysonrcecomponentsinthehigh—dimensionalsignalsweredistinguishedrfomthe

stationarycomponentsbyusingtheSSA．Then，theselectednon—stationarycomponentwiththemaximumkurtosisvalue

wasde—noisedbytheminimumentropydeconvolution(MED)．Finally，thede—noisednon-stationarycomponentwas

analyzedbytheenvelopespecturmtoextractthefaultcharacteristicrfequency．Thesimulationandexperimentresults

indicatethattheproposedmethodcanextractthefaultrfequencybetterthantheenvelopedemodulationmethodbasedon

empiricalmodedecomposition(EMD)．

Keywords：phasespacereconsturction；stationarysubspaceanalysis；minimume