Neuro-Fuzzy Proportional-Integral-Differential Controller for Alternate Current Servomotor

This paper proposes a simple gain scheduling robust proportional-integral-differential (PID) controller whose parameters are continuously changed by fuzzy logic such that if the error e is large (if the S/N ratio is large), then the controller makes the system respond quickly and, if its rate e is large (a sudden disturbance is applied), then the integrator gain is very large and vice versa in order to obtain a robust PID controller that is insensitive to both plant noise and observation noise. The controller parameters are determined by using Liapunov stability theorem. The controller is tuned with a radial basis neural network with “off-line”, since a neuro chip is not commercially available. Then, it is practically applied to speed control for an induction servomotor. Experiments show that the proposed controller is superior to both a sliding mode controller and a PID controller.