Abstract：In view of that BP neural network has a slow convergence problem and poor reliability for a hoist braking-system fault diagnosis, a neural network fault diagnosis method based on improved particle swarm optimization algorithm is proposed.Because of the nonlinear and coupling between the fault symptoms and the fault of a braking system, a hoist braking-system fault diagnosis model is established.Improved particle swarm optimization algorithm is applied to optimize the weights and thresholds of the network, and the optimized neural network is used to diagnose braking system faults, which can shorten the neural network training time and improve the fault-diagnosis precision.The simulation results indicate that the diagnosis strategy has the characteristic of strong ability and high deficiency for diagnosis.

Keyword：mine hoist;braking system;fault diagnosis;neural network;particle swarm optimization algorithm;