A method for constructing Hedge Algebraic Type-2 Fuzzy Logic Systems
Phan Anh Phong, Đinh Khắc Đông, Tran Dinh Khang · 2011
In this paper, we have proposed a method for constructing Hedge Algebraic Type-2 Fuzzy Logic Systems (HaT2-FLS) from input-output data with two main phases. In the first phase, the data-driven Type-1 Fuzzy Logic System (T1-FLS) is designed under the combination of Fuzzy C-Means algorithm and Genetic Algorithm (GA). Then we construct the HaT2-FLS on the basis of the above T1-FLS in the second phase. The rule base of T1-FLS is employed for generating the rule base of HaT2-FLS by using the same number of fuzzy sets and rules in both fuzzy logic systems. The difference is that the antecedent and consequent sets of HaT2FLS themselves are Hedge Algebraic Type-2 Fuzzy Sets (HaT2FSs). In later phase, the fuzzy parameters of HaT2FSs are also optimized by GA. Finally, an example is provided to evaluate the performance of the proposed method.