Improved Fuzzy Expert System And Its Application In The Field Of Medical Diagnostic

Expert system is one of the most active fields of artificial intelligence, and it haswide applications. An expert system is an intelligent application software system,which contains a large number of human experts’ experiences and knowledge of thefield, and can use these to deal with the problems of this field. Fuzzy expert system isone kind of important expert systems.This paper mainly studies the rule-based fuzzy expert systems, focuses on themethods of knowledge representation and fuzzy reasoning, and applies the theoreticalresults to construct a disease diagnosis fuzzy expert system.In this thesis, research work and results are as follows:1) Different knowledge representation forms used in fuzzy expert systems arediscussed. The main advantages and disadvantages of these methods are reviewed,and the production rule form is selected for the rule-based fuzzy expert systems.2) Fuzzy reasoning is an important part of fuzzy expert systems, and it determinesperformance of a fuzzy expert system in large extent. CRI method is one of thefrequently used Reasoning methods, but it has some disadvantages. This shows thatthe reasoning method should be improved. Therefore, the literatures proposed manymethods of fuzzy reasoning, but these methods mostly confined to theoretical researchresults, which have not been applied to fuzzy expert systems. So this paper selected asuitable reasoning method for building fuzzy expert systems.3) Fuzzy expert systems have been applied to many professional fields. This paperconsidered the medical field as the research object. Knowledge of disease is acquiredthrough the investigations in various ways, and is converted to the rules which meetthese system requirements by the production rule representation method.4) A fuzzy expert system for disease diagnosis is constructed, implemented and simulated. The selected methods of knowledge representation and fuzzy reasoning areused to construct the medical diagnosis system, in which expert knowledge of thefield is integrated into the rule-based system. By using this system patients’ symptomsare diagnosed, and the corresponding results are reported. Simulation results showthat, the fuzzy expert system which is established in this paper can accomplish theautomatic diagnosis of26common diseases in the region, and the diagnosis resultsshow a higher diagnostic accuracy than others.