基于经验模式分解与小波软阈值的联合去噪法

PDF全文下载：2015040463

林助军， 严志军*， 肖敏， 朱新河， 程东(大连海事大学 轮机工程学院，辽宁 大连 116026）

摘要： 为获得往复运动摩擦副摩擦力信号良好的去噪效果，在分析传统的经验模式分解去噪和小波阈值去噪的优缺点的基础上，提出在模式相关系数准则下，将经验模式分解去噪与小波软阈值去噪结合起来的综合去噪方法。首先利用经验模式分解出若干个固有模式函数分量，在模式相关系数准则下找出由噪声主导和由信号主导的固有模式函数分量的分界点，接着对由信号主导的模式函数分量分别进行小波软阈值去噪，然后再进行重构从而获得去噪信号。实验结果表明，此去噪方法具有良好的自适应性和稳定性，提高了信噪比，并较好地保留了原信号的细节信息。

关键词： 经验模式分解； 小波软阈值； 相关系数； 去噪

中图分类号： TN 911.7文献标志码： A

 Signal Denoising Based on Empirical Mode Decomposition and Wavelet Soft Threshold

 LIN Zhujun, YAN Zhijun, XIAO Min, ZHU Xinhe, CHENG Dong

(School of Marine Engineering, Dalian Maritime University, Dalian 116026, China)

Abstract: In order to get a good denoising performance for reciprocating friction signals,a new comprehensive denoising approach based on empirical mode decomposition (EMD) and wavelet soft threshold denoising was proposed under the criterion of mode correlation coefficient.Firstly,the signal was decomposed into several intrinsic mode function components by using empirical mode decomposition.The dividing point between the noisedominated components and the signaldominated components was found out adaptively according to the criterion of mode correlation coefficient.Then the signaldominated components which were denoised respectively by using wavelet soft threshold were reconstructed to achieve denoising signal.The experimental results show that this denoising method has good adaptability and stability, improves the signal to noise ratio (SNR),and retains details of the original signal effectively.

Key words: empirical mode decomposition; wavelet soft threshold; correlation coefficient; denoising

收稿日期： 20140518

作者简介： 林助军(1990—)，男，硕士研究生. *通信联系人

文章编号：16726987（2015）04046305； DOI： 10.16351/j.16726987.2015.04.021