Fault diagnosis of bearing using wavelet packet transform and PSO-DV based neural network

In this paper, a fault diagnosis system is proposed for rolling bearing using wavelet packet transform (WPT), particle swarm optimization (PSO) algorithm with differential operator named PSO-DV and back-propagation neural network (BPNN) techniques. In the preprocessing of vibration signals, WPT coefficients are used for evaluating their entropy and treated as the features to distinguish the fault conditions of bearing. In the classification, to verify the effect of the proposed PSO-DV based BPNN in fault diagnosis of bearing, a classical PSO based BPNN is compared with a PSO-DV based BPNN. The experimental results showed the proposed intelligent method can escape from local minima, so has better convergence and diagnosis ability than classical PSO based BPNN. Meanwhile, it achieves classification of bearing fault.