In view of the nonlinear mapping ability of artificial neural network, the ability of self-learning and adaptive uncertainty system dynamic characteristic, fault tolerant and generalization ability and parallel processing ability, etc., the article puts forward a PID adaptive control algorithm based on neural network inverse as well, introducing RBF neural network to the inverse control, and operating PID integration. In a sudden external disturbance and model parameter change, the control scheme can significantly reduce resistance perturbation influence on speed, and strong robustness on parameter variations and external disturbances of the system.