基于小波包和包络分析的滚动轴承故障自动诊断方法  

Rolling bearings fault auto-diagnosis based on wavelet packet and envelope analysis

作　　者：贾军峰[1] 杨国安[1] 李新华 吴振生[1] 

机构地区：[1]北京化工大学故障与自愈工程研究中心,北京100029 [2]新疆泽普塔西南石油勘探开发公司,新疆泽普844804

出　　处：《石油矿场机械》2006年第5期18-22,共5页Oil Field Equipment

基　　金：国家自然科学基金资助项目(50575017)

摘　　要：提出了一种基于峭度-小波包-包络分析的滚动轴承故障自动诊断方法。首先用峭度系数判断是否出现故障,再针对故障轴承振动信号非平稳和调幅的特点,用小波包将信号分解到不同的节点上。然后将不同节点的重构信号做包络谱分析,将谱峰处的频率同滚动轴承内圈、外圈、滚珠、保持架等的故障频率进行对比分析,在此基础上,运用最小距离判别法则自动获得故障原因,从而实现滚动轴承故障的自动诊断。通过对实验中采集到的轴承振动信号进行分析,证明了该方法在轴承早期故障诊断中的有效性。A method of auto-diagnosis based on kurtosis, wavelet packet and envelope is presented for the rolling bearing fault diagnosis. The kurtosis is employed to analyze if the fault has occurred. The fault signals which are non-stationary and modulated are decomposed at different levels by the wavelet packet. The envelope spectrum of reconstructed signals at different levels is illustrated. By contrasting the frequency on spectrum peak and the fault frequency of inner race, outer race, balls and rolling element, the auto-diagnosis of the rolling bearing fault is realized by min-distance principle. Satisfactory results have been obtained from the experimental vibration signals and the effectiveness of the proposed method is demonstrated in auto-diagnosing for the rolling bearing faults.

关 键 词：峭度 小波包 包络 滚动轴承 自动诊断 

分 类 号：TE9[石油与天然气工程—石油机械设备]