A generalised fuzzy reasoning algorithm for an object‐oriented expert system tool
暂无分享,去创建一个
: A fuzzy reasoning algorithm is presented in this paper. It is based on the concept of a generalised fuzzy production rule that adds new conjunctions such as ADD and REL to the usual AND and OR conjunctions for the linkage of premises in conventional rules. A definition is given for weights and related factors, to express and measure a relation between the premises. A formal representation for fuzzy premises is presented, together with the introduction of a fuzzy match method. Multiple thresholds are used to convert the uncertainty measures of reasoning results into linguistic variables. We have applied the fuzzy reasoning algorithm to a frame-based hybrid expert system tool. An object-oriented approach is used in implementing the system. Some results of the application are presented and discussed.
[1] R. Ramakrishnan,et al. The fuzzy weighted additive rule , 1992 .
[2] Janusz Kacprzyk. Compatibility relations for the representation of associations between variables in knowledge-based systems, and their use in approximate reasoning , 1991 .
[3] H. E. Stephanou,et al. Perspectives on imperfect information processing , 1987, IEEE Transactions on Systems, Man, and Cybernetics.
[4] Zhang Yue,et al. A fuzzy production rule based expert system , 1991 .