Fault diagnosis of belt conveyor gearbox based on WT feature-enhanced cICA

Constrained independent component analysis (cICA) algorithm has a strong denosing ability for measured noise mixed in measured signals,but is very poor for source signal with source noise. To overcome this problem,a method of gearbox fault feature extraction based on wavelet transform(WT) feature-enhanced cICA is proposed. It can improve signal-to-noise ratio (SNR) and enhance the analysis effect of cICA algorithm with the wavelet decomposition of measured signal and reconstruction for a certain sub-band wavelet coefficients. However,the analysis effect of cICA method is not good without WT denosing. The results of simulation analysis and fault diagnosis of mine belt conveyor gearbox show that the proposed method can reduce the influence of source noise,and extract the gear fault feature,es- pecially weak low-frequency fault feature. It provides a new effective approach and mean for the multi-channel vibra- tion condition monitoring and fault diagnosis of mine gearbox.