Research and application of radial basis network bogie fault diagnosis model based on particle swarm optimization

Bogie system is the key system that affects the safety and quality of EMU operation. The construction of fault diagnosis model for bogie system can effectively improve the safety and comfort of EMU operation. The traditional modeling method uses BP neural network to model by fitting bogie system temperature and other parameters. However, BP neural network is prone to fall into local minimum, slow convergence and poor diagnostic accuracy. In this paper, particle radial basis function neural network (PSRB) is designed by using particle swarm optimization algorithm with high convergence. Particle Swarm optimization (PSO) is used to optimize the parameters of RBF Neural Networks. According to the complexity of the input parameters of the bogie system, the input and output parameters of the model are determined. Particle swarm optimization algorithm is used to search the optimal values of the center, width and output layer weight threshold of the RBF neural network. The hybrid algorithm is applied to the fault diagnosis of bogie system, and a bogie fault diagnosis model based on particle radial basis function neural network is designed. The experimental results show that the diagnosis model can effectively improve the identification accuracy of fault diagnosis, the minimum error accuracy is 0.0055, the operation time is saved, the operation time is reduced to 1.9s, and the influence of non-target parameters on the inversion results is eliminated. The model can also be used in other EMU systems, and has practical application value.