Research on fault diagnosis method of bearing based on parameter optimization VMD and improved DBN

Abstract

Aiming at the problem that the bearing characteristics are difficult to extract accurately, and the fault diagnosis is difficult. This paper proposed a novel bearing fault diagnosis method with parameter optimization variational mode decomposition (VMD) and particle swarm optimization Deep Belief Networks (PSO-DBN). Firstly, the PSO is applied to optimize the parameter of the VMD and solve the problem of parameter setting of the VMD. Then, to effectively extract the feature information, using the optimized VMD, the original signal is decomposed into intrinsic mode components, and each component's dispersion entropy (DE) value is calculated. Finally, to further improve the accuracy of fault diagnosis, the PSO-DBN model is used to recognize the fault pattern bearing. The results of both experiments are 100 %. The results show that this method can effectively extract bearing fault features and accurately realize fault diagnosis. Compared with other methods, the accuracy of this method is increased by at least 2.08 % and the maximum is increased by 33.33 %.