Self-Tuning PID Controller Based on Improved BP Neural Network

In order to solve the difficult problem that how to reduce the overshot and shorten the regulating time of the PID controller based on BP neural network, a self-tuning PID controller based on improved BP neural network is presented. The parameters of the PID controller are calculated by an improved BP Neural Network according to the input and output and the error of the PID controller. It is introduced the dynamic adjustment for activation function in the output layer, and the dynamic adjustment for learning rate to improve the Fletcher-Reeves conjugate gradient method. In the simulation experiments in the Matlab 7.0, two plants are selected to test the performance of the proposed PID controller, and also the rand noise are added to the input to test the robustness of them. From the simulation results, the overshoot are lower than those of controllers by using the steepest descent method and the Fletcher-Reeves conjugate gradient method; the regulating time is also shorter than those of controllers by using the steepest descent method and the Fletcher-Reeves conjugate gradient method; the proposed control algorithm is more robust than those of controllers by using the steepest descent method and the Fletcher-Reeves conjugate gradient method.