Fault diagnosis of automobile rear axle based on Wavelet Packet and Support Vector Machine

This paper presents a fault diagnosis method of automobile rear axle based on wavelet packet analysis (WPA) and support vector machine (SVM) classifier. By Fourier transformation we find out the frequency band that can mostly reflect the rear axle failure state and use wavelet packet to decompose and reconstruct the vibration signals of rear axle, then extract each band's energy and the variance, standard deviation, skewness, kurtosis of the specific frequency band to constitute a feature vector. We use the feature vectors which are come from some pieces of normal and abnormal samples to train support vector machine classifier for obtaining the best classification, at the same time, discuss the optimization of SVM parameters. Application shows that the method is effective in real time fault diagnosis for the automobile rear axle and has a strong anti-interference ability in different working conditions.