结合复小波包变换及频谱校正的机械转子碰摩故障诊断方法研究

针对机械转子系统中碰摩故障发生时故障特征微弱及识别困难的问题,提出一种结合双树复小波包变换及频谱校正的故障诊断方法。首先对于振动位移信号中工频基波成分,采用频谱加矩形窗的频谱校正方法识别其谐波信息,通过构造补偿信号进行对消,以减少其对后续特征提取的影响。其次通过双树复小波包对补偿过的信号进行多尺度分解;最后对小波包子空间信号进行希尔伯特包络解调分析,通过瞬时幅值及瞬时频率信息诊断转子的动静碰摩故障。在转子实验台上进行了实验验证,结果表明提出的方法能有效提取转子碰摩产生的微弱故障特征。

Research on Rubbing Fault Diagnosis of Mechanical Rotor Based on Complex Wavelet Packet Transform and Spectrum Correction

In order to solve the problem of weak fault characteristics and difficulty in identification when rubbing faults occurred in mechanical rotor systems, a fault diagnosis method combining double-tree complex wavelet packet transformation and spectrum correction was proposed. Firstly, for the fundamental wave component of vibration displacement signal, the harmonic information was identified by using the method of frequency spectrum plus rectangular window, and the compensation signal was constructed to cancel, so as to reduce its influence on the subsequent feature extraction. Secondly, the multi-scale decomposition of the compensated signal was carried out by using the double-tree complex wavelet packet. Finally, the Hilbert envelope demodulation analysis of wavelet steamed bun spatial signal was carried out to diagnose the rotor dynamic and static rubbing fault by instantaneous amplitude and frequency information.The results show that the proposed method can be effectively used to extract the weak fault characteristics of rotor rubbing.