The application of fuzzy control to a coke oven gas cooling plant

Abstract This paper describes a fuzzy control system for a gas cooling plant using a special-purpose fuzzy inference accelerator. The plant described had been manually controlled for over twenty years, with several failed attempts at automation. Long time lags, constantly changing system parameters, and uneven gas flow rates are some of the problems that our control system has successfully addressed. Actual working results from the plant's one and a half years of fuzzy control are presented that demonstrate the control system's reliability.