Design And Implement Of Fuzzy Expert System Based On Predicate

Expert System is an important branch of Artificial Intelligence with wide application during practical fields. There are variable presentations for knowledge in Expert System. However, during practical applications, the performance of the knowledge presentation is always reflected directly by the efficiency of the accomplishment of tasks. Therefore, based on original fuzzy Expert System,this paper digs into the combination of predicate presentation and fuzzy inference, which greatly enhanced the knowledge presentation ability of the original system. So a new Expert System with more powerful presentation ability and reasoning ability is built up.This Expert System is composed of fuzzy inference machine and management of fuzzy knowledge base. The fuzzy inference machine is based on forward data-driven reasoning and the fuzzy knowledge base is based on predicate presentation. This paper develops a reasoning algorithm special for the combination of predicate presentation and fuzzy knowledge, which effectively solves the conflict between predicate logic and uncertain reasoning.The knowledge bases system in this expert system is self-contained system, which not only can store the accurate numerical knowledge and the fuzzy language expression knowledge accurately and conveniently, but also fuzzy continuous knowledge and dispersed knowledge effectively. It solved the problem of the store of subjection function and domain in the database, not concerning the type of the database. The knowledge base system is made up of rule base, predicate base, variable base and dictionary base. Management and maintenance of the fuzzy dynamic database of knowledge is operated in visual interference, and realized in reasoning machine with new method special for combination of predicate presentation and fuzzy knowledge.The effectiveness of this expert system has been proven during tests, which supplies a gap between the predicate logic and uncertain inference.