基于小波包分解的滚动轴承故障诊断  

Fault Diagnosis of Rolling Bearing Based on Wavelet Packet Decomposition

作　　者：黄中华 尹泽勇[2] 刘少军[1] 丁文强[2] 

机构地区：[1]中南大学机电工程学院,湖南长沙410083 [2]中国航空动力机械研究所,湖南株洲412002

出　　处：《湖南科技大学学报（自然科学版）》2008年第2期32-35,共4页Journal of Hunan University of Science And Technology：Natural Science Edition

基　　金：国家自然科学基金资助项目(50675226);湖南省自然科学基金资助项目(08JJ4011)

摘　　要：提出了一种基于小波包分析的滚动轴承故障诊断方法用于实现滚动轴承早期故障的检测.该方法的诊断过程如下:对轴承原始振动信号进行频谱分析,获取振动信号能量集中的频段.根据频段的范围和振动信号的采样频率确定小波包分解的层数.采用小波包分解的方法提取滚动轴承振动信号中能量集中的频段并生成相应的重构信号,对重构后的振动信号进行Hilbert变换和二次频谱分析.通过对比轴承故障的特征频率和二次频谱中的特征谱线判断轴承是否有故障及其发生位置.运用上述方法对具有外环故障的滚动轴承进行了实验研究并成功地实现了滚动轴承外环故障的检测.实验结果表明基于小波包分析的诊断方法可以有效诊断出滚动轴承的早期故障.A method of fault diagnosis of rolling bearing based on wavelet packet decomposition was proposed to diagnose incipient fault of rolling bearing. The procedure of this method was following, got the frequency range of energy concentration area with spectrum analysis of initial vibration signal of rolling bearing, the number of decomposition of wavelet packet was determined by the frequency range of energy concentration area and the sampling frequency. Extracted the signal with wavelet packet decomposition and reconstructure the signal, which was in the range of energy concentration area. Hilbert change and secondary spectrum analysis were used to deal with the reconstructured signal sequently. With the comparison of characteristic frequency of wiling bearing fault and that of secondary spectral analysis, the location of fault can be obtained. Experiment were carried out to validate the performance of this method, which used a rolling bearing with incipient fault on outer raceway. Experiment results show that this method is feasible if incipent fault diagnose of wiling bearing. 6figs., 15refs.

关 键 词：轴承故障诊断 小波包分析 HILBERT变换 频谱分析 

分 类 号：TH133.3[机械工程—机械制造及自动化]