A Fuzzy Rule Interpreter to Build Expert Systems Based on Fuzzy Logic. An Application in Company Diagnosis

Abstract

In this paper we present Cognos, a general purpose tool for interpreting fuzzy logic rules. The inference engine works with intersection and union operators and in the defuzzyfication module the centroid method is provided. A confidence degree, measuring the goodness of the result based on the distribution of the obtained fuzzy set, has been introduced. Due to the high level of modularity of the tool, new defuzzyfication functions and operators can be easily added. Fuzzy sets are defined using piece-wise linear functions that give a good flexibility degree in the specification of the membership functions. We illustrate the usage of the designed language with a set of rules incorporated in the application that Logic Control has developed for company diagnosis based on this interpreter and included in its Manager Vision package.