一种基于小波的奇异信号检测方法  

A Method of Singular Signal Detecting Based on Wavelet

作　　者：孙娟[1] 唐慧强[1] 

机构地区：[1]南京气象学院信息工程系,南京210044

出　　处：《武汉理工大学学报》2004年第8期84-86,89,共4页Journal of Wuhan University of Technology

基　　金：江苏省攻关计划 (BE2 0 0 2 32 7)

摘　　要：介绍了小波二进制变换的基本原理 ,简述了小波变换奇异性和信号突变的关系。基于小波变换 ,给出一种结合3σ准则、软硬阈值折衷法的奇异信号小波检测方法。仿真结果表明此法既能有效地消除噪声 ,又能较好保留奇异信号 ,从而改善信号中奇异性信息的提取效果 ;在对信号奇异程度和奇异点位置的确定上 ,较以往的处理方法有明显的改善。The singular points and irregular transient parts of the singular signal are the embodies of important signal which is one of the main characteristics of a singular signal. The principle of wavelet binary system transform was introduced and the relationship between singularity of wavelet transformation and acute signal was expatiated. Based on wavelet, there is a new method of singular signal deteting which combines 3σ-rule and compromise model ( between the hard-thresholding and soft-thresholding models). The experimental results demonstrate that this method is better in signal filtering and acute signal keeping, which improves the abstract effect of the singular information of the signal. The method is better than some old ones in confirming the level of the singular of the signal and the location of the singular points.

关 键 词：小波分析 软硬阈值折衷法 奇异性 3δ准则 

分 类 号：TN911.7[电子电信—通信与信息系统]