滚动轴承故障诊断的改进小波变换谱峭度法  

Improved Spectral Kurtosis Algorithm Based on Wavelet Transformation for Rolling Bearing Fault Diagnosis

作　　者：沈金伟[1] 石林锁[1] 

机构地区：[1]第二炮兵工程学院五系,西安710025

出　　处：《轴承》2010年第8期46-49,共4页Bearing

摘　　要：在Morlet小波变换谱峭度法的基础上做了相应改进,利用小波变换谱峭度法得到的谱峭度值,自动构造最优匹配滤波器。将此方法应用于滚动轴承故障诊断中,并与原小波谱峭度法进行比较。结果表明,基于小波变换的谱峭度法的效果对滤波器的选取比较敏感,具有更加优良的监测和诊断效果。Based on Morlet wavelet transformation,the algorithm of spectral kurtosis（SK） is improved appropriately.The calculated results of SK are used by the new method to construct an optimum match filter.The method is applied in fault diagnosis of rolling bearings and compared with the former algorithm.The results indicate that the present algorithm is sensitive to the selection of filters and demonstrates better results in surveillance and diagnosis.

关 键 词：滚动轴承 故障诊断 Morlet小波变换 谱峭度 

分 类 号：TH133.33[机械工程—机械制造及自动化] TN911.7[电子电信—通信与信息系统]