Aimingat the problem that RBF neural network have a slow convergence speed and the diagnostic accuracy is low when applied to fault diagnosis, a fault diagnosis method of mine ventilator based on Adaptive genetic algorithm optimized RBF neural network is proposed.Using adaptive genetic algorithm to optimize the number of hidden layer nodes,the center and width of hidden layer function, and the generalization ability of network is improved.Through a large number of collection and finishing work to form a sample set,using the training sample set to train the network,make fault diagnosis of mine ventilator according to the network output results.The simulation reveals that compared with RBF neural network, the RBF neural network optimized by adaptive genetic algorithm has a faster convergence speed and less number of iterations. It can effectively identify the type of failure, and it has a higher accuracy of fault diagnosis.