Aimed at the unpredictability of the mechanical equipment wear, RBF neural network prediction model was put forward, and combined with particle swarm optimization algorithm (PSOA) to optimize the model parameters. The gear wear experimental inputoutput data of the transmission model for SG1352 series K727840ZW was used to establish the output prediction model by RBF neural network based on PSOA, and then compared with the predication of traditional AR model, BP neural network model and Hermite neural network model. The simulation results show that the RBF neural network model based on PSOA has simpler structure and higher prediction precision, and the validity and practicability of the proposed method is verified.