Fault Diagnosis for Rotator in Rotating Machinery Based on Support Vector Machine with Particle Swarm Optimization Algorithm

Abstract

As failure of rotator in rotating machinery has a certain concealment, fault diagnosis for rotator in rotating machinery based on support vector machine with particle swarm optimization algorithm is presented in the paper. And particle swarm optimization algorithm is applied to select the suitable parameters of support vector machine. In the study, we employ three PSO-SVM classifiers to recognize the four states of rotator in rotating machinery including normal state, rotor imbalance, rotor winding and rotor misalignment. More than 70 cases are used to testify the effectiveness of the PSO-SVM model compared with other classification models. The experimental results show that diagnostic precision for rotating machinery of PSO-SVM than that of SVM and BPNN.