基于小波变换模极大值的信号奇异性检测

Singularity Detection of Signals Based on Wavelet Transform Modulus Maximum

信号的突变性或奇异性通常携带有最重要的信息,小波变换为信号的多尺度瞬态分析提供了最合适的方法,本文阐述了基于小波变换模极大值的奇异信号检测理论,利用从光滑函数的一阶或二阶导数获得的小波,分析了信号奇异点定位方法及奇异度的计算方法,并指出了利用此方法时对所用小波函数的要求及尺度参数的选取原则,为非平稳信号的检测提供一种行之有效的方法。

The points of sharp variation or singularities usually carry the important information about the signals.The advent of wavelet transform provides a suitable framework for studying the multiscale transient representation of signals.This paper introduces the theory of singularity detection based on wavelet transformation modulus maximum.By wavelest that are the first or second derivatives of a smoothing function,analysis the methods about locating singularity points and computing their intensity.Based on these results,the choice of wavelet bases and scales parameters are studied.The simulation shows this metheod is very effective.