A fuzzy-logic based expert system for cupola furnaces

Abstract

Successful operation of a cupola furnace is highly dependent upon the mitigation of operational problems occurring in them. Cupolas rely on the experience of the operator for a smooth Operation. The focus of this research is to develop a fuzq-logic based expert system, which would provide expert advice to a cupola operator to address multiple operational problems occurring simultaneously in the cupola. An expert system for a cupola operation is not currently available. Expert sub-systems, which address individual operational problems, are deveIoped and tested These sub-systems are integrated using an intelligent methodology so as to provide suggestions that address multiple operational problems simultaneously. A serf-corrective capabili@ was introduced into the expert system by developing an iterative process for some vital parameters in the cupola furnace. Provisions were given to the user to prioritize the inputs and seek the pertinent suggestions fiom expert system. The expert sysfem was tested on a set of operational problems. It performed satisfactorily. As expected some problems could not be addressed simultaneously as they require contradictory solutions. The user was given the ability to modify the requirements so that a solution satisfiing part of the requirements can be achieved