Fault diagnosis of rolling bearing vibration based on particle swarm optimization-RBF neural network

The training procedures of RBF neural network are faster than BP neural network and it has the global optimal ability. However, a key problem by using the RBF neural network approach is about how to choose the optimal the parameters of RBF neural network. Particle swarm optimization is introduced to select the parameters of RBF neural network. In the paper, particle swarm optimization and RBF neural network method is applied to fault diagnosis of rolling bearing. Finally, the result of fault diagnosis cases shows high classification diagnostic accuracy in fault diagnosis of rolling bearing.