EXPERIMENTS WITH THE USE OF A RULE-BASED SELF-ORGANIZING CONTROLLER FOR ROBOTICS APPLICATIONS

The Rule-Based Self-Organising Controller (SOC) has as one of its main advantages the fact that there is no need to have a mathematical description of the system to be controlled. In this controller, the rules are linguistic statements expressed mathematically through the concepts of Fuzzy Sets and correspond to the actions a human operator would take when controlling a given process. This paper describes several experiments which were performed with SOC. The control was considered of both simulated linear and non-linear processes. The performance of the controller was investigated for step inputs as well as for waveforms which the process output had to track; the effect of process noise was also considered. It was shown that SOC had a performance slightly superior to that of a PID controller when noise was present and the process had time varying coefficients as well as pronounced non-linear characteristics. This paper also reviews some of the successful initial work carried out in connection with the application of SOC to a revolute joint robot arm.