Type-II T-S Fuzzy Modeling And Control

This article aims at the problem of handling the influence from uncertainties, researches the Type-II fuzzy modeling and the control method based on the Type-II fuzzy model. Compare with the crisp number membership of type-I fuzzy set, the membership of type-II fuzzy set is a type-I fuzzy set, the Fuzziness is enhanced, the ability of uncertainties description is improved, so type-II fuzzy system has the unique strength to handle uncertainties. T-S fuzzy model combines the strength of fuzzy system and linear system, it can express the fuzzy relationship by segmented linear polynomials, and it is convenient to combine with the linear control algorithms, so this article mainly studies algorithm.A new method for Type-II T-S fuzzy modeling is proposed in this article, first,the data samples are divided into several Type-I fuzzy sets by G-K cluster algorithm, then the Type-I fuzzy sets are extended to Type-II sets according to the analysis of the data samples to get the antecedent parameters of Type-II T-S fuzzy model, afterwards, we analyse the influence from uncertainties on the output of data samples to extend the consequent parameters from crisp number to Type-I fuzzy set, finally, we get the Type-II T-S fuzzy model. This method can get a High Accuracy model by low computational complexity and few steps.On the basis of Type-II T-S fuzzy model, we use predictive control algorithm to design two types of controller: Controller 1 and Controller 2. Controller 1 is based on the relation formula of input and output, it can conveniently quote the existed method of Type-I fuzzy control. Controller 2 use the method of extending Type-I to Type-II in the controller design, a control fuzzy set is computed, we can get the optimal control by defuzzing the control fuzzy set, in this way, the ability of the controller to handle uncertainties is improved, meanwhile the computational complex is increased.