Dynamic Knowledge Inference and Learning of Fuzzy Petri Net Expert System Based on Self-Adaptation Learning Techniques

Abstract

It is rather limited for fuzzy production rules to describe the vague and modified knowledge of expert system, an automatic fuzzy reasoning and learning framework based on fuzzy Petri net are presented for design a dynamic expert knowledge system in this paper. Fuzzy Petri net may describe the relative degree of each proposition in the antecedent contributing to the consequent accurately. In order to reason and learn expediently, FPN without loop is transformed into hierarchy model and continuous functions to approximate transition firing and fuzzy reasoning. The self-adaptation learning techniques based on back-propagation are used to learn and train parameters of fuzzy production rules of FPN. Simulation experiment shows that the improved adaptive learning techniques can make rule parameters obtain optimal or at least nearly optimal convergence rapidly.