A framework for object-oriented fuzzy expert systems

摘要

In recent years, the fuzzy logic has been widely used in the fieldof AI. Many successful fuzzy expert systems and fuzzy control systemshave been developed. In general, these systems adopt the rule-basedknowledge representation. On the other hand, it is also suitable toadopt the object-oriented paradigm for representing knowledge. Thispaper proposes a framework that integrates the fuzzy inference model andthe object-oriented paradigm for building expert systems. This frameworkembeds the properties of object-oriented paradigm, e.g., encapsulation,inheritance, message passing and polymorphism, into the model of fuzzyexpert systems. It also includes a partial inheritance model, so thatthe child object may inherits the property of parent object partially.The inheritance process is viewed as a process of evolution, and thepartial inheritance is treated as a special kind of fuzzy implication