预览加载中,请您耐心等待几秒...
1/2
2/2

在线预览结束,喜欢就下载吧,查找使用更方便

如果您无法下载资料,请参考说明:

1、部分资料下载需要金币,请确保您的账户上有足够的金币

2、已购买过的文档,再次下载不重复扣费

3、资料包下载后请先用软件解压,在使用对应软件打开

一种改进的自适应Chirplet分解方法及其实现方法(英文) Introduction Chirpletdecompositionisapowerfultoolforanalyzingnon-stationarysignals,suchasthoseencounteredinseismicdataprocessing,speechrecognition,andimageprocessing.However,theconventionalchirpletdecompositionalgorithmsuffersfromsomelimitations,suchassensitivitytonoiseandparameterselection.Inthispaper,weproposeanimprovedadaptivechirpletdecompositionalgorithmandanimplementationmethodtoovercometheseshortcomings. ImprovedAdaptiveChirpletDecompositionAlgorithm Theproposedalgorithmisanenhancementoftheconventionalchirpletdecompositionalgorithm,whichdecomposesasignalintoasumofchirplets,eachconsistingofacarrierfrequency,amplitude,andtime-varyingfrequencymodulation.Theimprovedalgorithmadoptsanadaptiveapproachtoselectthenumberofchirplets,adjusttheirparameters,andfilterthedecompositionresultstoreducetheimpactofnoise. Theadaptiveapproachisbasedonthefollowingsteps: Step1:Decomposethesignalusingtheconventionalchirpletdecompositionalgorithmtoobtainaninitialdecomposition. Step2:Calculatetheresidualoftheinitialdecompositionbysubtractingthesumofthechirpletsfromtheoriginalsignal. Step3:Setathresholdfortheresidualandcheckifitisbelowthethreshold.Ifso,outputthechirpletdecomposition;otherwise,proceedtoStep4. Step4:Selectasubsetofthechirpletsthathavesignificantamplitudesandmergethemintoanewchirpletusingaleast-squaresmethod.AddthenewchirplettothesumofchirpletsandrepeatfromStep2. Theadaptiveapproachtakesadvantageofthesparsityofthechirpletrepresentation,whichmeansthatonlyasmallnumberofchirpletsareneededtorepresentthesignalaccurately.Theresidualthresholdisusedtocontrolthenumberofchirplets,andthemergingprocessisusedtoimprovetheaccuracyoftherepresentationbyreducingtheimpactofnoise. ImplementationMethod Theproposedalgorithmcanbeimplementedusingthefollowingsteps: Step1:Selectawindowsizethatcoversasufficientdurationofthesignalandasuitablewaveletfortime-frequencylocalization. Step2:Applythewavelettransformtothesignalandobtainthescalogram. Step3:Selectascalerangeandper