| Expert system is a programme system provided with a substantial expertise and experience. According to a field of knowledge and experience provided by one or more human experts, it applies artificial intelligence technology to solve those complex problems which need the expert to decide. Aiming at overcoming the limitations of traditional rule-representation method of expert knowledge-base, This thesis studies the fusion technology of expert system and Fuzzy Petri Net (FPN), develops the FPN modelling and reasoning tool, and realizes the core mechnism of expert system.This thesis firstly introduces the knowledge representation and reasoning mechnism of FPN, proposes the mapping relation between production rules and FPN, gives the first order predicate logic representation algorithm of FPN which combined the logic inference and knowledge representation of FPN, and proves the inference in logic. Based on this, the definition of FPN basic elements in Visual Prolog is put forward to realize the expert system knowledge representation and reasoning based on FPN. This thesis sets forth the design plan of expert system based on FPN, gives the overall frame, discusses each function module, and mainly studies the knowledge-base expression method and its maintenance algorithm.According to the proposed design plan, this thesis further developes the FPN modelling and reasoning tool in Visual Prolog. This tool provides the graphical edit environment used to build the FPN model, constructs the FPN reasoning mechnism inside, explains the reasoning results to users through the explanation mechnism, and thereby realizes the development of expert system based on FPN.Finally, a case study on the TE process diagnosis is implemented by the FPN modelling and reasoning tool. The case shows that the FPN modelling and reasoning tool can establish the knowledge-base of expert system conveniently and effectly, and realize the reasoning, having the broad development foreground. |
PDF Full Text Request