動態(tài)測試非平穩(wěn)信號的系統(tǒng)參數(shù)識別方法研究.pdf

1、南京航空航天大學碩士學位論文動態(tài)測試非平穩(wěn)信號的系統(tǒng)參數(shù)識別方法研究姓名：周祥申請學位級別：碩士專業(yè)：機械設計及理論指導教師：岳林20100101動態(tài)測試非平穩(wěn)信號的系統(tǒng)參數(shù)識別方法研究AbstractTestsignalprocessinginbackgroundofbignoiseandshortsamplesisanallimportanttopicinthefieldofEngineeringapplications，andth

2、esignalofteninvolvelowerqualityofthetestsignaldata，thedataareoftenmodal—intensive，nonlineartheeffectivesampleisshortthenoiseratioofdatasignalislowandtheprocessofnonstationarycharacteristicsSo，inthispaperstudyandrealizean

3、ewtime—frequencydomainfilterdenoisingmethodandmodalparameteridentificationmethodstobettersolvetheengineemagrelatedissuesMaincontentsofthisarticleareasfollows：1Toenhancenoiseratiooftestsignal，proposedalinearsweepingsignal

4、denoisingmethodwhichisbasedonFractionalFourierTransformtotheImpatientalgorithm，asageneralizedformofthetraditionalFouriertransform，therearemanytraditionalFouriertransformfeaturesarenotavailableThealgorithmusesthesweepsign

5、alofthefocusingpropertiesoftimefrequencydomain，usingcovereveryotheroperationinthefractionalFourierdomaintoachievethepurposeofSignaltonoiseseparation，Comparedwimthetraditionalfilteringmethod，thismethodisbetterandCalculati

6、onissmaller2InordertobettersolvetestmodalparameteridentificationproblemoflowaccuracyproposedmultireferencepointLeast—squarescomplexfrequencydomainmethod(poIy_LCSF)，applicablebothtotheweakdampingalsoappliestostrongmodalda

7、mpingintensivesystemsThealgorithmusesalinearleast—squares—likealgorithmisestimatedthattheoverallmodelcoefficients，avoidingthedesignandoptimizationofcomplexnonlineardependenceontheinitialvalue，andsuccessfullyappliedtoexpe

8、rimentaldataanalysis，toachievebetterresults3Basedontheaboveprincipleofalgorithmsandothersimilarmethods，usingGUIinterfacetoolsinMATLABtodevelopasimpleandpracticalanalysissoftwareIncludingthevibrationsignalpreprocessingmod

9、ule，time—frequencyprocessingmoduleandtimefrequencydomainmodalparameteridentificationmoduleAfterverification，thesoftwarecanbeeasilyappliedtovoice，andvibrations(suchastheflightfluttertestsweepsignal)avarietyofsignalanalysi