An adaptive IMC-PID control scheme based on neural network

An adaptive IMC-PID control method based on neural network (NN) is proposed for the typical industrial process in this paper. The conventional IMC-PID provides a convenient method of tuning parameter according to the requirement of the control performance for closed-loop system because it has only one regulable parameter. But when the parameters variation and uncertainty factors are included in the control system, the controller parameter of IMC-PID should be re-tuned on-line. Therefore, the proposed method is applied to adjust the parameter of IMC-PID controller through the self learning of neural network, so as to enhance the robustness and control performance of the system. The weights of the NN are adjusted by the back propagation arithmetic so that the control error can be minimized. Simulation results show that the proposed method could achieve a better system performance than the conventional IMC-PID does.