Tunable Q-factor wavelet transform denoising with neighboring coefficients and its application to rotating machinery fault diagnosis

Abstract

Fault diagnosis of rotating machinery is of great importance to the high quality products and long-term safe operation. However, the useful weak features are usually corrupted by strong background noise, thus increasing the difficulty of the feature extraction. Thereby, a novel denoising method based on the tunable Q-factor wavelet transform (TQWT) using neighboring coefficients is proposed in this article. The emerging TQWT possesses excellent properties compared with the conventional constant-Q wavelet transforms, which can tune Q-factor according to the oscillatory behavior of the signal. Meanwhile, neighboring coefficients denoising is adopted to avoid the overkill of conventional term-by-term thresholding techniques. Because of having the combined advantages of the two methods, the presented denoising method is more practical and effective than other methods. The proposed method is applied to a simulated signal, a rolling element bearing with an outer race defect from antenna transmission chain and a gearbox fault detection case. The processing results demonstrate that the proposed method can successfully identify the fault features, showing that this method is more effective than the conventional wavelet thresholding denoising methods, term-by-term TQWT denoising schemes and spectral kurtosis.