Weighted Fuzzy Reasoning Based On Marked Fuzzy Petri Net

The marked fuzzy Petri net (MFPN) model is presented to represent the weighted fuzzy reasoning (WFR) and the corresponding algorithm is offered in this paper. The token in MFPN is endowed with the new meaning which is the fuzzy set but not the real number in [0,1]. According to the change of the token value and the algorithm presented in this paper, the final goal of the weighted fuzzy reasoning can be calculated. In the process of weighted fuzzy reasoning using the marked fuzzy Petri net, several knowledge representation parameters: local weight and the threshold value assigned to the antecedent proposition of the weighted fuzzy production rule, certainty factor of the deduced consequence etc. are considered in this paper. To make the reasoning suit the practice and increase the precision of the reasoning, the certainty factor of the weighted fuzzy production rule and the fact input by user is represented by use of fuzzy numbers. Based on the work before, fuzzy Petri net in this paper is applied in a general process of reasoning method that makes fuzzy process realize more easily.