An intelligent PID controller on RBFNN

To improve the quality control of applications and to address the problem of traditional PID controller´ parameters need to rely on trial and expert experience, the neural network PID controller is constructed, which combines the neural network theory with the PID control algorithm. The controller realizes automatic parameter adjustment on line. In the simulation system, the neural network PID controller based on the Gaussian kernel function not only can tune the PID parameters, but also effectively improve the convergence rate and control accuracy.