基于EEMD消噪和相关系数识别的滚动轴承故障诊断方法.docx

基于EEMD消噪和相关系数识别的滚动轴承故障诊断方法
摘要
随着机械设备的广泛使用，滚动轴承的故障诊断已成为一项重要的研究领域。本文提出了一种基于经验模态分解（EEMD）消噪和相关系数识别的滚动轴承故障诊断方法，该方法首先采用EEMD算法对滚动轴承信号进行分解和去噪，然后利用相关系数和相关系数矩阵进行滚动轴承故障诊断。实验结果表明，该方法能够有效地识别不同类型的滚动轴承故障，并取得了较高的诊断准确率和鲁棒性。
关键词：经验模态分解、相关系数、滚动轴承、故障诊断、消噪
Abstract
Rollingbearingfaultdiagnosishasbecomeanimportantresearchfieldwiththewidespreaduseofmechanicalequipment.Inthispaper,arollingbearingfaultdiagnosismethodbasedonempiricalmodedecomposition(EEMD)denoisingandcorrelationcoefficientrecognitionisproposed.Firstly,theEEMDalgorithmisusedtodecomposeanddenoisetherollingbearingsignal,andthenthecorrelationcoefficientandcorrelationcoefficientmatrixareusedforrollingbearingfaultdiagnosis.Theexperimentalresultsshowthattheproposedmethodcaneffectivelyidentifydifferenttypesofrollingbearingfaultsandachievehighdiagnosticaccuracyandrobustness.
Keywords:empiricalmodedecomposition,correlationcoefficient,rollingbearing,faultdiagnosis,denoising
1.Introduction
Withthedevelopmentofmechanicalequipment,rollingbearingsarewidelyusedinvariousmachineryandequipment,andthereliabilityofrollingbearingsdirectlyaffectsthesafeandstableoperationofequipment.Therefore,itisimportanttodiagnosethefaultsofrollingbearingsintimeandtakecorrespondingmeasurestoavoidseriousconsequences.Traditionalfaultdiagnosismethodsmainlyincludevibrationanalysis,acousticemissionanalysis,andtemperaturemeasurement.However,thesemethodsoftenrequirecomplexinstrumentationandprofessionalknowledge,andtheiraccuracyandrobustnessarenotveryhigh.
Empiricalmodedecomposition(EMD)isadata-drivenmethodthatcanbeusedtodecomposenon-stationarysignalsintoasetofintrinsicmodefunctions(IMFs)andaresiduecomponent,whichiswidelyusedinsignalprocessing.However,theEMDmethodhassomedrawbacks,suchasmodemixingandboundaryeffects.Toovercomethesedrawbacks,Huangetal.proposedtheEEMDmethod,whichaddswhitenoisetotheoriginalsignaltoreducemodemixingandimprovedecompositionaccuracy.
Correlationcoefficientisameasureofthedegreeoflinearrelationshipbetweentwovariables,anditiswidelyusedinstatisticalan