自适应Morlet小波降噪方法及在轴承故障特征提取中的应用  

Denoising method based on adaptive Morlet wavelet and its application in rolling bearing fault feature extraction

作　　者：蒋永华[1] 汤宝平[1] 董绍江[1] 

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：《仪器仪表学报》2010年第12期2712-2717,共6页Chinese Journal of Scientific Instrument

基　　金：国家863计划(2009AA04Z411);国家自然科学基金(50875272);高等学校博士点基金(20090191110005)资助项目

摘　　要：分析了Morlet小波变换的滤波特性及其时频分辨率,利用Morlet小波良好的时域和频域特性及奇异值分解技术,提出了一种基于自适应Morlet小波和SVD的降噪方法。针对滚动轴承故障在振动信号中表现为冲击衰减波形的特点,采用修正的Shannon熵方法同时优化Morlet小波的中心频率与带宽参数,实现其与冲击特征成分的最优匹配;针对根据小波系数矩阵奇异值曲线的过渡阶段求取最佳变换尺度的方法存在着不够快捷方便的不足,将其与小波系数奇异值比方法相结合来快速方便地求得最佳变换尺度;最后对信号进行降噪处理提取故障特征。对仿真信号和实际轴承内外圈故障信号的应用分析表明,该方法具有良好的降噪性能,能有效地提取出滚动轴承的微弱故障特征。The filtering characteristics and time-frequency resolution of Morlet wavelet transform are analyzed.Using excellent characteristics of Morlet wavelet in time domain and frequency domain and singular value decomposition technology,a denoising method based on adaptive Morlet wavelet and SVD is proposed.According to the fact that rolling bearing defect can excite vibration waveform with specific impact component,modified Shannon wavelet entropy is used to optimize the central frequency and bandwidth parameter of the Morlet wavelet to achieve optimal match with the impact component.Then,the method based on the transitional segment of singular curve of wavelet coefficient matrix combined with the method based on singular value ratio is used to obtain the optimal scale for the wavelet transform rapidly and conveniently.Finally,the signal is denoised and the fault feature is extracted.Experimental results and analysis results for rolling bearing signals with inner-race and outer-race faults show that the proposed method is feasible and effective for extracting weak fault feature.

关 键 词：MORLET小波 奇异值分解 奇异值比 微弱特征提取 

分 类 号：TH165.3[机械工程—机械制造及自动化] TN911.2[电子电信—通信与信息系统]