A neural network robust controller for real-time control of induction motor

A neural network robust controller (NNRC) to enhance the robustness of the conventional feedback controller is proposed in this paper. The control system consists of an optimal conventional feedback controller and a neural network robust controller in parallel to the controlled system. The optimal controller is used to guarantee the stability of the whole control system and present the optimal tracking performance to the reference input signal. To overcome the uncertainties included in the plant, a neural network robust controller with an on-line learning algorithm is added to the conventional optimal controller for the identiÞcation of uncertainties and adaptive modiÞed control at the same time. The new controller has increased robustness with regard to parameter variations. The new controller also exhibits very good disturbance rejection property. Simulation experiments and real-time control are made to illustrate the quality and robustness of control achieved for the induction motor system.