A systematic approach to obtaining fuzzy sets for control systems

The behavior of a control practitioner, who may be acting only on subjective or intuitive knowledge could be represented by one or more fuzzy subsets of the set (assumed to be finite) of all states of a controlled system. In this paper a systematic approach is proposed for determining the grades of membership of such fuzzy sets. These sets and the plant are then optimized against a performance index. The results of applying this method to a particular second order system are presented and then compared with those obtained by using a continuous deterministic model of the same plant. Implications for the design of controllers are then drawn.>