Fault Diagnosis Method for Vehicle-Mounted Hybrid Excitation Motor Drive System

In this paper, a method to perform the fault diagnosis of the drive system of vehicle-mounted hybrid excitation synchronous motors, combining particle swarm optimization and chimpanzee optimization algorithm (PSO-ChOA) with variational mode decomposition (VMD), approximate entropy for fault feature extraction and extreme learning machine (ELM) is proposed. First, a model of a six-phase hybrid excitation synchronous motor is constructed and six typical faults are simulated. Second, the classical variational mode decomposition algorithm is used to extract the feature vectors of the 7 states (including normal state) of the model from the fault signals. Finally, in order to effectively extract the features of the six-phase motor signals and perform diagnosis, the proposed chimpanzee optimization algorithm is used to optimize the key parameters of variational mode decomposition and extreme learning machine, respectively, which improves the accuracy of six-phase motor fault diagnosis. The experimental analysis results show that the fault identification accuracy of the method reaches 99.27%, which can identify motor faults more effectively than traditional algorithms.