基于小波包分析的滚动轴承故障特征提取  

Feature Extraction in Fault Diagnosis of Rolling Element Bearing Based on Wavelet Packet

作　　者：陈季云[1] 陈晓平[1] 

机构地区：[1]江苏大学电气信息工程学院,江苏镇江212013

出　　处：《微计算机信息》2007年第02S期192-193,219,共3页Control & Automation

基　　金：国家自然科学基金(No.30370395)

摘　　要：简述了小波包分析的基本原理及其用于特征提取的机理,利用小波包对滚动轴承振动加速度信号进行分解,求出各频率段的能量,并以此作为滚动轴承所发生故障的特征向量进行提取,从而识别出滚动轴承的故障,通过对于实测信号的分析证明了该方法的有效性,体现了小波包分析的优良性。The mathematical principle of wavelet packet analysis and its application to feature extraction for fault diagnosis are introduced in this paper. The wavelet packets is firstly used to transform the transient vibration signals into frequency domain , and then the energy in different frequency band is extracted to use as features vector of fault bearing. It is confirmed by the experiment on a test equipment that this method has some advantages and can be applied in fault diagnose of rolling bearings successfully, which indicates the advantage of wavelet packet analysis.

关 键 词：小波包分析 故障诊断 滚动轴承 

分 类 号：TH165.3[机械工程—机械制造及自动化]