ROBUST DESIGN OF RULE-BASED FUZZY CONTROLLERS

This paper deals with the design of stable and robust rule-based fuzzy control systems. New expressions to compute indices which provide a m easure of the stability and robustness of fuzzy control systems are pr esented. The relation between the modification of the rules and the st ability is studied through the so-called sensitivity indices. The pape r presents procedures that make use of these indices to improve the de sign of fuzzy control systems, including the modification of the rules to obtain the global stability of an unstable system with multiple at tractors, and to improve the dynamic behavior or the robustness of a n on-linear plant. An example with a fuzzy controller for a system with non-linear damping and saturation in the actuation is presented to ill ustrate the design procedure.