Use of causal knowledge in a real-time fuzzy logic controller

The establishment of expert control techniques and their actual industrial impact have been inhibited by the lack of a comprehensive design methodology. This paper summarizes principal drawbacks limiting the development of a definitive methodology. It also addresses the key problem of acquisition of plant knowledge, clarifying first the types of knowledge. For a case study, a real-time fuzzy-rule-based controller using causal knowledge is designed for control of liquid level in a pilot plant and tested for closed-loop operation under different conditions of set-point change, noise disturbance and parametric change. The results show that it yields superior performance over a qualitative and a conventional proportional integral derivative controller implemented on the same plant and subjected to similar disturbances.