Bearing fault diagnosis of rolling bearing based on weighted-permutation entropy and DE-ELM

Aiming at the feature that vibration signal of rolling bearing is non-stationary and non-linear, a bearing fault diagnosis method of rolling bearing based on weighted-permutation entropy and differential evolution algorithm(DE) optimized extreme learning machine(DE-ELM) is proposed. Firstly, complete ensemble empirical mode decomposition with adaptive noise was done for vibration signal of rolling bearing to obtain several intrinsic mode functions (IMFs). Then, weighted-permutation entropy of main IMFs was calculated as feature vector of fault signal. Finally, differential evolution algorithm (DE) was used to optimize input weights and hidden layer bias of extreme learning machine, and feature vector of fault signal was taken as input of DE-ELM. The experiment results show that weighted-permutation entropy can accurately extract fault features, and DE-ELM algorithm can effectively improve the accuracy of fault diagnosis. This method is more accurate and reliable compared with several methods.