Machine condition monitoring and fault diagnosis based on support vector machine

Due to the importance of rotating machinery as one of the most widely used industrial element, development a proper monitoring and fault diagnosis technique to prevent malfunction and failure of machine during operation is necessary. This paper presents a method for gearbox fault diagnosis based on feature extraction technique, distance evaluation technique and the support vector machines (SVMs) ensemble. The method consists of three stages. Firstly, the features of raw data are extracted through the wavelet packet transform (WPT) and time-domain statistical features. Secondly, the compensation distance evaluation technique is applied to select optimal feature via sensitivities ranking. Finally, the optimal features are input into the SVMs to identify different faults. The diagnosis result shows that the SVMs ensemble is able to reliable recognize not only different faults styles and severities but also the compound faults in high accurate rate.