Fault diagnosis of wind power gearbox based on IPSO-FNN

Wind power gearbox (WPG) is an important component of wind turbines, which is prone to faults due to factors such as working environment and complex structure. Furthermore, the causality of faults is difficult to be expressed by the mathematical model. A fault diagnosis method of WPG is proposed based on the improved particle swarm optimization combined with fuzzy neural network (IPSO-FNN). In order to improve the efficiency of network learning algorithm, IPSO is used to learn network parameters. Fitness variance is introduced to characterize the state of particles, and differential evolution is applied to premature particles to improve the diversity of particle swarm. Simulation results show that the proposed method has higher diagnostic accuracy and faster convergence speed than FNN and PSO-FNN methods.