Skip to Main Content
The application of a self-organizing fuzzy proportional-integral-derivative (SOF-PID) controller to a multiple-input-multiple-output (MIMO) nonlinear revolute-joint robot arm is studied in this paper. The SOF controller is a learning supervisory controller, making small changes to the values of the PID gains while the system is in operation. In effect, the SOF controller replaces an experienced human operator, which otherwise would have readjusted the PID gains during the system operation. The three PID gains are tuned using classical tuning techniques prior to the application of the SOF controller at the supervisory level. Two trajectories of step input and path tracking were applied to the SOF-PID controller at the setpoint. For comparison purposes, the same experiments were repeated by using the self-organizing fuzzy controller (SOFC) and the PID controller subject to the same information supplied at the setpoint. For the step input, the SOF-PID controller produced a aster rise time, a smaller steady state error, and an insignificant overshoot than the SOFC and the PID controller. For the path tracking experiments, better results were obtained.