Fuzzy Expert System Research And Realization

Fuzzy expert system is an importance practice field front of knowledge engineering. It is a key research content of expert system to carry on effective management, maintenance of fuzzy knowledge and study, realization on its uncertain reasoning technology. Base on the national technological department project-Orange Diseases Diagnosing System, the paper has researched the design and realization of reasoning engineer based on fuzzy technology.In the performance of expert system, the indefinite information and the fuzzy information in real world caused the expert system get the limit, in order to effectively simulate these information, after massive research, I mainly used the definite theoretical model in the inference process to infer. The definite theoretical model is the indefinite inference method model that united the theory of probability based on the reliability method, the merit of definite theoretical model was the using direct and the effect is good. The definite theoretical model mainly divides into three parts: indefinite description to knowledge, indefinite description to as well as definite factor renewal algorithm. Moreover, I choose reverse infers control strategy in the system.We use the relational database to replace the knowledge base, use common programming language to compile inferring engine, refers to the work way of prolog engine, the citrus fruits expert system inference engine, mainly includes the design of blackboard, the design of fuzzy inducing equipment as well as the design of interpreter equipment.This research results of article bring forth as follows: The fuzzy mathematics and the fuzzy theory and its the practical application, the Prolog language blur research, by develop the fuzzy expert system, the express of fuzzy knowledge base using the relational database, the designs and research fuzzy of reasoning engine, including blackboard, inducing equipment and interpreter. But I have not solved all problems, there are following some key questions, the knowledge automatic gaining, the indication of knowledge uses a better method, with its correspondence inference engine design and so on.