Automation of diagnosis of electric power transformers in Itaipu Hydroelectric Plant with a fuzzy expert system

A method for automating the maintenance of electric power transformers by means of Fuzzy Expert System is presented. The rules and the Knowledge Base of that system have been obtained from the usual proceedings adopted by the Brazilian Itaipu Hydroelectric Plant, as well as from the Company´s expert personnel. The adoption of a fuzzy methodology in the determination of the diagnosis showed to be quite appropriate to approach the problem. In this paper, several details regarding the choice of the modular structure of the Expert System and its rules are discussed, including the implementation of the Key Gas criterion associated with Rogers criterion. The original version of that criterion has been extended in order to comprise a greater number alternatives originating from the relationships among the types of gases found dissolved in the insulating oil.