A fuzzy rule-based method of controlling dynamic processes

The first rule-based controller for industrial processes using fuzzy logic was proposed seven years ago. Many research teams have continued work on this system since then. However, because early work emphasised the use of fuzzy mathematics, the underlying mechanisms within the controller have tended to be obscured. The result is a tendency towards misdirected inquiry of some superficial features of the controller. This paper deals mainly with an intuitive understanding of the underlying process. A similar type of analysis is also extended to the self-organising controller in which the rules are learnt through experience.