小波去噪中噪声估计失真判别方法-噪声与振动控制.PDFVIP

第35 卷 第5 期 噪 声 与 振 动 控 制 Vol 35 No.5 2015 年10月 NOISE AND VIBRATION CONTROL Oct. 2015 文章编号：1006-1355(2015)05-0145-05+ 173 小波去噪中噪声估计失真判别方法 王 新，王建顺 （河南理工大学 电气工程与自动化学院，河南 焦作 454000 ） 摘 要：研究小波阈值去噪时存在的噪声估计失真问题。当采用常见的阈值确定方法对含有较强高频分量的信号 进行小波去噪时，小波分析的频带能量泄漏现象会导致噪声估计失真，从而使小波阈值去噪出现较大的偏差。从小波 分解的d 1细节层和d2 细节层的相关性角度，揭示d2 细节层频带能量泄漏对噪声估计影响的规律，提出根据d 1、d2 细 节层的最大相关系数判别噪声估计失真的方法。最后，给出解决噪声估计失真的方法。实验表明，该方法可以很好地 判别小波去噪中是否出现噪声估计失真，可以避免出现去噪后信号有用信息损失严重的问题。 关键词：声学；小波去噪；噪声估计失真；阈值方法；频带能量泄漏 中图分类号：TN911.7 文献标识码：A DOI 编码：10.3969/j.issn. 1006-1335.2015.05.030 Identification Method for Noise Estimation Distortion in Wavelet Denoising WANG Xin , WANG Jian-shun ( School of Electrical Engineering and Automation, Henan Polytechnic University, Jiaozuo 454000, Henan China ) Abstract : When the signals containing some high frequency components are denoised by using the wavelet denoising with conventional threshold selection methods, the frequency- band energy leakage in wavelet analysis will lead to the distortion of the noise estimation, so that the wavelet threshold denoising may contain large errors. In this paper, the noise estimation distortion in the wavelet threshold denoising was studied. In view of the relevance between the d1 detail level and the d2 detail level, the influence of the frequency-band energy leakage of the d2 detail level on the noise estimation was analyzed. The identification method for noise estimation distortion was proposed by using the maximum correlation coefficient of the d1 detail level and the d2 detail level. Finally,