Fault Diagnosis Based on Principal Component Analysis and Support Vector Machine for Rolling Element Bearings

摘要

An intelligent fault diagnosis method of rolling element bearings based on statistics analysis, principal component analysis (PCA), and support vector machine (SVM) is presented. The method consists of three stages. First, due to the fact that it is hard to obtain sufficient fault samples in practice, different features are extracted as many as possible to acquire more fault characteristic information. Second, the original feature set is compressed with PCA from 43 to 4 dimensions. Finally, the compressed feature set is fed into SVM classifier to identify the fault patterns of the rolling element bearings. For comparison, the back propagation neural network (BPNN) is also utilized to solve the same problem. The results show that the proposed method can achieve higher accuracy and adaptability than BPNN when facing high-dimensional, nonlinear, and a smaller number of samples.