Application of Neural Network in Optimization of PID Controller

Optimization of PID controller parameters has been a hot issue in the fields of Automatic control. In the automatic control process, the controlled object has nonlinear and uncertainty characteristics. Traditional PID parameters methods are often time-consuming and difficult to obtain control effect, causing the control accuracy not high. In order to solve the optimization problem of PID controller parameters and improve system performance, we propose optimization method of PID parameters based on neural network. This method regards PID controller as the original input of neural network, the optimal parameters as the output of neural network. PID control parameters are dynamically adjusted in the control process to optimize itself by the associative memory of neural network and self-learning. Simulation results compared with the traditional PID parameters optimization method show that, this method has strong robustness and improves the system response speed, its anti-interference ability and adapt to the changing of parameter that is superior to the conventional PID control.