Self-organising fuzzy logic control and the selection of its scaling factors

Fuzzy logic control has been applied successfully to many industrial processes. It is particularly appropriate when process models are either unknown, non-linear, or variable in structure. There are two problems in its usage. One is the obtaining of a suitable rule-base for the application, while the other is the selection of scale factors prior to fuzzification and after defuzzification. Self-organising fuzzy logic control obviates the first problem, and is demonstrated in this paper. The selection of the scale factors is more difficult, but heuristic rules are derived in this paper via experimentations on a wide range of processes. The rules have been elicited from single variable control of a liquid-level rig, an air-heating process, an electrical-drive controller, and an anaesthesia model. These heuristic rules are then generalised to the multi-variable case of a coupled-electric drives system and simultaneous control of muscle relaxation and unconsciousness in anaesthesia.