Takagi-Sugeno Intuitionsitic Fuzzy Inference Systems Based on Game Theory

We describe in this paper three approach for calculating intuitionistic index in design a Takagi-Sugeno intuitionistic fuzzy inference system(TSIFIS). The new approaches are based on game theory. The expression of output of TSIFIS is analyzed, and is split into two parts: agreement part P1 and non-agreement P2. There is a matrix game between agreement part P1 and non-agreement P2. Based on the intuitionistic rule, the intuitionistic index matrix is established. Maxmin, minmax and average strategies are present to calculate the intuitionistic index matrix using the intuitionistic index matrix. Finally a typical example has indicated that the approaches to calculate intuitionistic index is a rational and effective methodlogy and the TSIFIS based on new approaches for calculating intuitionistic index provides stronger possibility to accommodate imprecise information compared to fuzzy inference systems.