基于最优Morlet小波和自项窗的混合时频分析方法研究  

Hybrid time-frequency method based on optimal Morlet wavelet and auto terms window

作　　者：刘文艺[1] 汤宝平[1] 陈仁祥 

机构地区：[1]重庆大学机械传动国家重点实验室,重庆400030

出　　处：《振动与冲击》2010年第9期5-8,27,共5页Journal of Vibration and Shock

基　　金：国家高技术研究发展计划(863计划)(2009AA04Z411);国家自然科学基金(50875272);霍英东教育基金会11届青年教师基金(111057)

摘　　要：提出一种基于最优Morlet小波和自项窗的混合时频分析方法。对和机械冲击信号波形相似度较高的Morlet小波进行改进,采用交叉验证法和Shannon熵方法设计了改进Morlet小波参数和小波变换尺度,对信号进行连续小波变换(CWT)以实现滤波消噪;然后,设计了自适应自项窗函数,对Wigner-Ville分布(WVD)交叉项进行移除,消除WVD交叉项的干扰。仿真和实验验证了所提出的方法可以有效地对含噪信号进行滤波消噪、并去除WVD中干扰项的影响,提高时频分析的分辨率和能量聚集性。A new hybrid time-frequency method was put forward based on the optimal Morlet wavelet and auto terms window.The Morlet wavelet,with a shape similar to the mechanical shock signals,was improved by adding two parameters to determine the shape of the mother wavelet.The added parameters and the appropriate scale parameter for the continuous wavelet transformation were designed by the cross validation method and the minimum shannon entropy method.Then,the useful components of the signal analyzed were obtained by the improved Morlet wavelet de-noising method.The auto terms window based on the smoothed pseudo Wigner-Ville distribution(SPWVD) spectrum was used as a window function to remove the cross terms in Wigner-Ville distribution(WVD).The simulation and fault diagnosis experiment results show that the proposed method has a good de-nosing performance and is effective in the cross terms removing and fault feature extraction.

关 键 词：小波消噪 MORLET小波 自项窗 混合时频分析 Wigner-Ville分布(WVD) 

分 类 号：TN911.7[电子电信—通信与信息系统]