Application of improved wavelet packet energy entropy and GA-SVM in rolling bearing fault diagnosis

In view of the problem that the feature vectors are difficult to extract accurately in the mechanical fault diagnosis, taking rolling bearing fault as an example, a new fault diagnosis method based on the combination of the improved wavelet packet energy entropy and the GA-SVM(GA optimization SVM algorithm) classification algorithm is proposed. The improved wavelet packet analysis method is used to decompose the collected signals by multi-layer wavelet packet, reconstructed decomposition signal by single branch, extracting the wavelet packet energy entropy, formation of feature vectors for fault diagnosis, and using it as input to establish fault diagnosis model for GA optimized SVM to realize the status recognition of the rolling bearing. The experimental results show that this method has higher classification accuracy than the unmodified wavelet packet energy entropy, which can improve the accuracy of the state recognition of rolling bearings and effectively achieve the fault diagnosis of rolling bearings.