基于EMD的包絡(luò)相關(guān)算法研究.pdf

1、基于EMD的包絡(luò)相關(guān)算法研究專業(yè)名稱：電子科學(xué)與技術(shù)申請人：何經(jīng)偉指導(dǎo)教師：胡維平教授論文答辯委員會主席：委員：蘭露廣西師范大學(xué)碩士論文ResearchonthemethodrelatetotheenvelopeofEMDMasterCandidate：HeJingweiSupervisor：HuWeipingMajor：ElectronicscienceandtechnologyResearchDirection：signalproc

2、essingGrade：2012AbstractAlmosta11thesignaliSnonlinearnonstationaryinreallifeThetimefrequencyanalysismethodhasbeenpaymoreattentiontostudyforprocessingnon1inearandnonstationarysignalHilberthuangtransformisanewmethodoftimef

3、requencyanalysisafterFouriertransfoITn，wavelettransformThecoreideaofhilberthuangtransformiStheempiricalmodedecompositionEmpiricalModeDecomposition(EMD)isalocaldigitalsignalprocessingmethod，fullybasedonadaptivedata，iSvery

4、suitableforanalyzingnonstationaryandnonlinearproperties，multicomponentsignalThemethodbreaksthroughtheconventionalideasoftraditionalsignalprocessing，theprocessingmethodissimple，andCanmoreaccurateanalysisinstantaneouschara

5、cteristicsofthenonlinearsignal111emethodhasgoodapplicationprospectinnonlinearnonstationarysignalprocessingdirection，thuswontheattentionofacademiccirclesAtpresent，manyscholarshascarriedontheimprovementofrelatedresearchinv

6、iewofthedefectsanddisadvantagesoftlleEMD，hasachievedgreatachievementmaketheflawsoftheEMDarewellimproved，hasprovidedabroadplatformfortheengineeringapplicationoftheEMD111iSpapermainlystudytheEMDtheorymethodThespecificworkc

7、ontentofthepaperasfollows：First，ThisPaperdetailelaboratedthestandardempiricalmodedecompositionalgorithmwhichbyNEHuangputforwardSecondaccordingtothestandardEMDhastheincompleteenvelopesituationwhenstructuretheenvelopeincub

8、icsplineinterpolationmethodnleproblemistheenvelopewillcuttingsignallineImprovetheenvelopealgorithmttoimprovethestandardsEMDalgorithm，andtheeffectivenessoftheimprovemethodiSverifiedbyexperimentTllirdtheoptimalmeanenvelope

9、，theorthogonalitystoppingcriteriaoftheinverseEMDalgorithmprocessWasintroducedinthisPaperTheauthortestandcomparedtheorthogonalityindexandenergy1eakindexofthestandardEMD，beforeoftheimprovedinverseEMD，andtheimprovedinverseE

10、MD，analysedandcomparedtheadvantagesofoptimalenvelopemeanandotherenvelopemean，studiedtheeffectofEMDdecompositionbyenhancesamplingrateFourth，thisP印erintroducesthefastbandpassfilterinwinsicmodefunctiondecompositionmethod『31

11、thatputforwardbylizhenzhangThefastfilteringmethodCaneffectivelyovercomethemodalaliasingphenomenonisverifiedbyanexample碭einnovationofthepaperasfollows：First，throughthepointoftangencytofittingenvelopeinsteadoftheconvention

12、almethodofextremevaluepointtofittingenvelopeintheenvelopefittingprocess，CanavoidappearringovershootandundershootphenomenonwhenfittingenvelopeneartheextremevaluepointtheeffectivenessoftheimproveingtheEMDalgorithmiSverifie

13、dbyexperimentSecond，thesecondorderoptimalenvelopemethodisproposedinreversescreeningoftheEMDalgorithmtoavoidexcessivescreeningouttheothercomponentwhichofwithoutphysicalmeaning，thispaperuseimprovingenvelopemethodintheEMDal