Font Size: a A A

Adaptive Time-frequency Analysis For Non-stationary Signal

Posted on:2013-01-17Degree:MasterType:Thesis
Country:ChinaCandidate:L L MaFull Text:PDF
GTID:2248330395457034Subject:Control Engineering
Abstract/Summary:PDF Full Text Request
Time-frequency analysis is hot in singnal processing.It provides effective tools fordealing with the non-stationary signals that are ofen encountered in natures.This paperpresents parametric adaptive time-frequency analysis method using Gauss linearfrequency modulation wavelet as the basis function and then dividing the signal intobasis function’s linear combination,on the basic of studying the typical lineartime-frequency analysis, quadratic time-frequency analysis and adaptive time-frequencyanalysis.Though simulation and error analysis,it shows that this method eliminates thecrossover effect and maintain high time-frequency aggregation.The first and second chapter introduce the non-stationary signal’s concept and themain mathematical tools.And it summaries domestic and international’s presentresearch situation, and the purpose and significance of the study.The third chapter introduces the non-stationary signal’s linear time-frequencyanalysis, such as Short time Fourier transform (STFT), Gabor extension and wavelettransform; quadratic time-frequency analysis, such as Wigner-ville (WVD) and Cohenclass.And it studies the relationship between them and their advantages anddisadvantages.The fourth chapter studies the adaptive time-frequency analysis method,includingadaptive kernel time-frequency analysis,time-frequency analysis based on adaptivefilter,parameter decomposition adaptive time-frequency analysis.Taking the radialGauss kernel for example,it simulates adaptive kernel time-frequency analysis’smethod and steps,as well as taking the instantaneous frequency measurement method fobroadband non-stationary signal fo example to simulate the adaptive filter’s applicationin time-frequency analysis.The paper focuses on parameter decomposition adaptivetime-frequency analysis.Though the basic function’s analysis,it selects Gauss linearfrequency modulation wavelet as the basis function,and then do decomposition of theanalysised signal and time-frequency analysis.After simulation,it proves that the methodis correct and effective.
Keywords/Search Tags:Signal processing, Time-frequency anslysis, Adaptive kernel, Adaptive filter, Parametric time-frequency analysis
PDF Full Text Request
Related items