Research on Fault Diagnosis of Rolling Bearing Based on Wavelet Packet Transform and IPSO-SVM

Abstract

For the difficulties of feature extraction of fault signals of rolling bearing and the limitation of structural parameter optimization of support vector machine(SVM), this paper proposes a method of fault feature extraction and classification based on wavelet packet transform and improved particle swarm optimization(IPSO)support vector machine. First, the feature is extracted using wavelet packet transform, and the sample entropy value of each band obtained by decomposition is used as the feature vector. Secondly, the IPSO algorithm is used to optimize the tow structural parameters of SVM, penalty and Gaussian kernel coefficients. Finally, a fault classification model for rolling bearing is established. Results showed that the fault diagnosis classification model based on wavelet packet transform and IPSO-SVM has higher accuracy.