关键词: 连续小波变换, 自相关系数, 齿轮, 故障诊断, 特征提取

Abstract:

In order to solve the problem that can not denoising the external noise when extracting fault feature of gear, this paper introduces a method that can identify the time of periodic impulsive fault signatures from the measured noisy signal mixture on the basis of CWT(Continuous Wavelet Transfon) and auto-correlation coefficient method. A comb filter can be applied to extract fault features in timescale domain, the spurious impulses can be removed effectively from the extracted fault feature. Experiments show that this method can accurately identifiy the fault feature of impulsive signals with missing tooth.

Key words: continuous wavelet transfon(CWT), autocorrelation coefficient, gear, fault diagnosis, feature extraction