Abstract
Fuzzy if-then Rules are used for knowledge representation in rule-based system. In order to model the overlapping areas among the rule sets corresponding to different classes, a new fuzzy reasoning algorithm based on fuzzy integral is proposed. Each of the rule sets is composed of the rules which have the same consequence. This algorithm introduces the Choquet integral and the possibility measure carries on the valid combination, thus transforms the question as the optimized model. Compared with the algorithm in [4], this algorithm can be better than it, can improve the reasoning accuracy and also strengthen the reasoning capability of rule-based system.