基于自适应频率窗经验小波变换的列车轮对轴承多故障诊断

Multiple Fault Diagnosis of Train Wheelset Bearing Based on Adaptive Frequency Window Empirical Wavelet Transform

针对列车轮对轴承故障信号复杂,尤其是在多故障并发情况下难以准确诊断的问题,提出了基于频率窗经验小波变换(EWT)的轮对轴承多故障诊断方法。首先对轴承多故障振动信号进行Fourier变换,引入一个带宽可变的滑动频率窗分割信号频谱;然后利用水循环优化算法(WCA),通过所提出的幅值包络谱相关峭度(ESCK)指标,自适应地确定轴承多故障中各单一故障所对应的最优频率窗位置;最后通过经验小波变换分解出单一故障信号,采用包络解调分析实现轴承复合故障准确诊断。轮对轴承多故障仿真和实际应用结果表明,所提方法能有效分离列车轮对轴承复合故障中的典型故障,有效降低轮对轴承多故障诊断的误诊率,具有一定的应用价值。

Aiming at the problem that wheelset bearing fault is difficult to diagnose under complicated fault vibration signals,especially in the case of concurrent multiple faults,a new multi-fault diagnosis method named self-adaptive frequency window empirical wavelet transform (EWT) for wheelset bearings was proposed in the paper. Firstly,a moving and flexible frequency window with variable bandwidth was introduced to segment the Fourier spectrum of signal,which is based on the Fourier transform of multi-fault vibration signals of the wheelset bearing. Secondly,the envelope spectrum correlated kurtosis (ESCK) combined with envelope demodulation and correlated kurtosis,as the evaluation criterion,was used to determine the optimal position of frequency window of each single fault in bearing multi-faults through the water cycle algorithm (WCA). Finally,the envelope demodulation analysis method was applied to process the mono-component signal induced by proposed method,and the single fault information was easily extracted. The simulation and practical application show that this method can separate single-channel multiple fault signals of wheelset bearing and further reduce misdiagnosis rate of train wheelset bearing fault diagnosis effectively,which has certain engineering application value.