Simulation Studies of an Analytical Fuzzy Tuner for a PID Servo

The performance of a fuzzy-logic tuner for a PID controller is studied. An analytical framework that has been developed elsewhere is used for this purpose. Specifically, the concept of resolution relation which links the modal values of a condition variable (an observed attribute of the system response) with those of an action variable (a tuning parameter increment), is employed. Two aspects of a resolution relation are the resolution offset and the nonlinearity. The simulation studies indicate that normally a linear resolution curve with zero offset is adequate for tuning a PID controller. Nevertheless, some potential for adaptively using these two aspects to improve the performance of a fuzzy-logic tuner is noted.