Real-time fuzzy expert control

This paper describes research into direct real-time fuzzy expert control. Two rule-based controllers, a Boolean and a fuzzy version, were designed to control various processes with minimal process knowledge. The rule base, which is exactly the same in both versions, uses two control strategies: direct expert control and direct reference expert control. Simulation as well as real-time results show that a wide range of processes can be controlled with little a priori information about the process dynamics. The fuzzy version of the rule base shows that fuzzification improves the performance of the rule-based controller before adaptation, but it is more time consuming. The rule-based controller can be adapted to the process to be controlled, which leads to a better control performance