An analytic approach to fuzzy robot control synthesis

The main advantage of a fuzzy control system is the fact that no mathematic al model of the controlled plant is required. Instead of that model, it is necessery to construct a fuzzy rule base for each particular application ca se. A vexing problem in fuzzy control, however, is the exponential growth i n rules as the number of variables increases. This problem is avoided here by the introduction of a new, nonconventional analytic method for synthesis ing the fuzzy control. For this purpose a new analytic function is defined that determines the positions of the centres of the output fuzzy sets, inst ead of the definition of a fuzzy rule base. This function can be adapted to each concrete application case by changing the free fuzzy-set parameters. The proposed analytic approach to the synthesis of fuzzy control, has been tested by a numerical simulation of an analytic fuzzy control system for a robot with four degrees of freedom. (C) 2000 Elsevier Science Ltd. All righ ts reserved.