发动机机械加工设备状态监测策略研究

上海交通大学硕士学位论文发动机机械加工设备状态监测策略研究姓名：李凌申请学位级别：硕士专业：车辆工程指导教师：何维廉;颜炳祥20071128IIIISTUDIESONENGINEMACHININGEQUIPMENTCONDITIONMONITORINGSTRATEGYABSTRACTThisdissertationresearchesenginemachiningequipmentthroughconditionmonitoringandfaultdiagnosis.Bystudyinghigh-speedmachiningcenterofit,IclarifytheFFTandwaveletcombinedtechnologyisagoodmeanstoestablishpredictivemaintenancesystem.Becauseenginemachiningequipmentsintheprocessingcenteristhekeypartsofanenginefactoryproductionline,thetimeperiodistolearnfaultsinthehighfrequencyspindle-theforemostpartofthemachiningcenter.Thentopreventsuddenmachinedownonthemachiningcenter,ensuresafetyinproduction,reducemaintenancecost,improvemanagementofequipment.Iintroducedtheideaofconditionmonitoringandfaultdiagnosis.AfterbriefingitsoriginanddevelopmentIelaborateFailureModes&EffectsAnalysisFEMA,whichincludesfaultmodes,faultcause,faulteffect,severityoffaultanddifficultyofdetectingfault.Typically,relatedknowledgeisneededinthisdissertation.BasedonvibrationsignalsIcangetinformationabouttheconditionofmachiningcenter,byusingspectrumanalysistechnology.Todothis,layoutsofaccelerationsenorandfeaturesofhighfrequencyspindlearealsoinvolved.Indeed,whatIstudiesisrotatingmachinerywhichhasitsinherentfaultmode.DatahavebeencollectedfromspindleontheteststandwhichIconstructfortheresearch.Iusethesedatatoexaminerelatedtheories.IVThecomputingresultsleadtomyconclusionusingFFTandwaveletanalysiscombinedtechnologycaneffectivelydiagnosefaultandlocateproblem.Bypropellingchangesinmaintenanceourcompanyhascutlotsofcosts.Apparently,conditionmonitoringandfaultdiagnosisprovidebasisforpredictivemaintenance.KEYWORDSwavelettransform,Fouriertransform,highfrequencyspindle,conditionmonitoring,faultdiagnosisVIIINr/minfiHzfoHzfcHzfbHzDmmdmmZ)(xϕ)(xψ20071128200711282007112811150/specialpurposemachine/TransfermachineTLTransferLinel2/HybridHSMC+SPM/TMAFTLAgileFlexibleTransferLineAFTL2“”“”12121[1]3U.K.MHMG&CMAR.A.Collacott,1978MechanicalFaultDiagnosisandConditionMonitoring[2]111[1][2]123---[1]124[1][1][2]1.2.3.4.5.122[1][3]5[2]1.;2.3.4.5010060220703758013,,,,3:1.SMDS,,,,,;,2.DMDS,SMDS,,DMDS,,,,6,,,3.RMDS90,Internet,--(RMDS),,,,;,3:patternclassificationrule-basedinference(mathematicalmodel)131,,,1.,,,,:FFT,2.,A1…AnUn,Uu0,A=max[A1(u0),…,An(u0)]u0Ai,:jAjp,Aj=(x1,…,xp),7xiAj,,Aj:)](),...,([1pxAjxAjAjµµ=XiXi(1.1):ij=Aj(xi)ixij,,,,,132,1-1,,:11Fig.11rules1W;2;3W;4=0)()(iixAjxAjµµ8F.Rosenblatt60,,90,,BP,,,,,100%;,,,133,,,ARMA1ARMAqkqkkpkpkkZZZ−−−−−−−=−−−ωθωθωφφ......11112ARkpkpkkZZZωφφ+++=−−...11(1.2)kZkωφφ=[1φ,…,pφ]θθ=[1θ,…,qθ]ARMAkZ3,,ARMAARMA(n,m)ARMA,,,ARMA;,ARMA;,ARMA;,:,,,914[4][5]50“”“”machineryhealthmonitoringMHM“”mobilediagnosiscenterMDCBK2500,:HP5451C,B&K2520BENTLY3300M6000B&K2520(vibrationmonitortype2520)623IRDIQ2000ROCKWELL-ENTEK10EMONITOR-ODYSSEYMAXIMO1986a.3MD-3MD-3MD-b.RMMDSc.RB20-1d.RMMDSe.CDMSf.CMD-IIIg.MD3905h.BJD-ZBJD-ZBJD-Z1985103MD-“”3MD-3MD-1987RMMD-S“”-1115Fourier1221FailureModes&EffectsAnalysisFMEAFMEA[3]123123RCM13123456FMEA22/122040m/min143CNCCNC(45)CNC32CPU64CPUNURBS(B)CNC(1)(2)(3)(4)4(PCD)(PCBN)(5)()21Fig.21HighFrequencySpindle12153323231110212112~34~56~78910162321102222233110232312~34~67~9101~23~45~67~89~10172411024267843871681472834784247561358120185402CNC18324324324183244165303612526224182006565025193160%-70%31131231Fig.31Vibrationclassificationf10Hzf=10-1000Hzf1000Hz203133df32Fig.32Threefeaturesofvibration2131433Fig.33VibrationSignalProcessing------------------------134Fig.34Waveimageofgearfailures22AB35Fig.35Autocorrelation2(3.1)36Fig.36Probabilitydensityfunctions()()[]∆=∆∆+≤≤=∞→→∆→∆TTxxxxtxxpxpxTxxlim1limlim00x(t)233,,37Fig.37Amplitudespectrum38Fig.38Cestrum24439Fig.39Phasespectrum5310Fig.310Three-dimensionalspectralarray25311Fig.311ZoomSpectrum32321)(txt∆)(tx)(nxT(3-12312Fig.312Sampling26)mfmf2/1Nyqusitmf23223-12bV1V1=100mV100mVA/D2B2=∆BbitB2562==∆BB=8)(nx02558bit8bit33FourierTransformedeFourier/mft21≤∆27,.,,1)(tf),(+∞−∞)(tfffπω2=2DFT)(tx[0,T]A/DN)(nxtetnxfXtfnjNn∆∆≈∆−−=∑π210)()(Ttnf/1/1=∆=∆NnjWe=−/2πn-n=0,1,2,...,N-1K-K=0,1,2,...,N-1DFT3FFTNW)(nx2N)1(−NN24N)1(4−NNN1965CooleyTukeyFFTNlog2N,N,FFT)(nxFFTFFT,nkNW:)()(NknNkNnNnkNWWW++==*)()(knNknNNWW=−*,∑−==10)()(NnnkNWnxkX∫+∞∞−−=dtetfFtjωω)()(∫+∞∞−−=dtetffFftjπ2)()((3.2)282FFT48FFTFFT4n22nnnbaA+=nfω3418221946GaborGaborSTFTSTFTSTFT)sincos(2)()(000tnbtnaaTntxtxnnωω++=+=)cos(2)(010nnntnAatxϕω++=∑∞=nnnab1tan−=ϕnϕnA(3.3)(3.4)29STFT(1))(tf)(tf)(0nfCn=∑−=nnntCtf)()(00ϕ=0101)(ttϕ∑−=nnntCtf)2()(11ϕ)(21012021++=nnnCCC)()()(101tftftg−=∑−=nnntdtg)2()(11ψ)(21012021+−=nnnCCd−=012112101)(tttψ302HaarϕψHaar)(tϕk)(kx−ϕ)(xϕk0VHaar{}),...22(),12(),2(),12(...,−−+xxxxnnnnϕϕϕϕnVn2/1nnnnVn2/1nVn2/1n2/1nW12/1−n1−nVnW)(xψ)(xϕnV)(xψnW)(xψ)(xϕnW∈=0)1,0[1)(xxϕRakxakZkk∈−∑∈),(ϕ31kanWnV1+nVnnnWVV⊕=+1)(2RL:)(2RLf∈3Haar)(tfnN2=Nyquistk)(tf)(tf)(tfn)()(...)()()(0021xfxWxWxWxfnnn++++=−−)2()(),2()(111111kxaxfkxbxWlZkZklklllkl−=−=−∈∈−−−−−∑∑ϕψ11−−lklkba∑∑∑∑∈−=∈−=−+−=+=ZkknlZkllknllnkxakxbxfxWxf)(])2([)()()(010010ϕψlkbRakxaknZkk∈−∑∈),2(ϕkkkkWwVfwff∈∈+=∑∞=,,0000...)(1002⊕⊕⊕