Research on a new fault diagnosis method based on WT, improved PSO and SVM for motor

Background: Motor is a device that converts electrical energy into mechanical energy. It is one of the most widely used mechanical equipments. Its running state directly affects the performance of machinery. Objective: The purpose of this study is to provide an overview about fault diagnosis methods from many literatures and patents, and propose a novel fault diagnosis method for ensuring safe operation of motor. Method: In this study, wavelet transform, improved particle swarm optimization algorithm and support vector machine are introduced into fault diagnosis to propose a novel fault diagnosis method. Wavelet transform decomposes the signal into different frequency bands. Then the fault features are effectively extracted by using frequency band analysis technique to be considered as the input vector of support vector machine. The strategies of dynamic linear adjustment, diversity mutation and adaptive inertia weight are used to improve particle swarm optimization algorithm, which is employed to optimize the parameters of support vector machine for obtaining a classifier with higher veracity. Results: The improved particle swarm optimization algorithm and proposed fault diagnosis method are fully evaluated by experiments and comparative studies. The results show that the proposed method takes on better classification accuracy and can quickly diagnose the motor faults. Conclusion: The improved particle swarm optimization algorithm has improved the convergence speed and precise. But because the motors often work in a noisy environment and recent patents on fault diagnosis method have also promoted new algorithm during their application. So the new fault diagnosis method with high accuracy is further studied in the future. © 2016 Bentham Science Publishers.