基于多小波分析的暂态信号去噪方法研究

由于多小波分析方法的特殊性,需要对原始信号进行预处理.不同的预处理方法对多小波去噪效果有着非常明显的影响.文中介绍了多小波理论,选用GHM多小波,采用了4种预处理方法对信号进行预处理,分析了基于这4种方法进行信号去噪的结果,并与单小波处理结果进行了比较.将比较结果应用于检测系统输出电压信号进行分析,结果表明,利用多小波变换方法对暂态信号去噪比用单小波方法具有明显的优越性.

Study of Transient Signal Denoising Based on Multi-wavelet Analysis

As a vital branch of wavelet research,multiwavelet is featured symmetry,orthogonality,and short support,which plays a great role in signal processing.Because of multi-wavelet's unique nature,preprocessing of original signals is required.Since different preprocessing methods have different effects on denoising for multi-wavelet,careful selection of preprocessing should be made.In the paper,first of all,Multi-wavelet theory is introduced.Then,with the use of GHM multi-wavelet,four methods are applied to preprocess signals.Next,their denoising effects are analyzed,followed a comparison with that of single wavelet.The comparison indicates,with properly selected preprocessing method,in terms of denoising,multi wavelet transformation is apparently superior to single wavelet.