On the application of fuzzy logic in the design of supervisors for real-time control systems

This work deals with the use of fuzzy logic techniques in order to include heuristic knowledge about the system under study into a supervisor module. The goal of the supervisor is to track the system operation between different states (such as different operation points, operation under hard nonlinearities, restricted zones, etc.). As the design is to be applied in real-time control system applications, the fuzzy supervisor is optimised in the sense of fast and reliable processing. The paper proposes a methodology for its design and shows its application on a servo-positioning system.