基于小波包频带能量特征的包络分析在滚动轴承故障诊断中的应用

经典的小波包迭代算法会由于小波包分解过程中的隔点采样而发生频率混叠现象。采用小波包移频算法，以1^#、2^#、3^#这3个406滚动轴承（其中1^#轴承工作正常，而2^#、3^#轴承工作状态未知）进行分析。通过小波包分解提取了这3个轴承振动信号的频带能量特征，初步判断2^#、3^#轴承的工作情况。基于此，再选择合适的频带进行小波包重构，并经Hilbert变换实现包络检波，这样轴承发生故障时的高频共振信号就实现了包络解调。再通过对振动包络信号进行频谱分析就可以对轴承的故障类型作出判定。

Wavelet Envelope Analysis Based on Wavelet Packet Energy Feature and Its Application in Failure Diagnosis of Rolling Bearing

A frequency-shift arithmetic is proposed in the text which overcomes the frequency aliasing of the classical wavelet packet transform and analyses three Type 406 rolling bearings of which,No.1 works well,how the others work aren't known. Their energy features are distilled by way of wavelet packet decomposition and based on which whether well the others work can be determined.Then the signals of relevant frequency bands are reconstructed and demodulated by way of Hilbert transform.Finally the failures are diagnosed by envelope spectrum analysis of the envelope signals.