A Modified Type-2 Fuzzy Logic System and Its Applications

Considering the complexity of designing an interval type-2 fuzzy logic system (IT2FLS) and the internal relationship between a type-l fuzzy logic system (T1FLS) and an IT2FLS, a modified type-2 fuzzy logic system learned by a T1FLS (MT2FLS) is proposed in this paper. The MT2FLS includes a T1FLS and an IT2FLS which is an extension of the T1FLS. The designing process of MT2FLS mainly includes two parts: fuzzy rule base learning and parameter learning. The initial rule base in the MT2FLS is empty. On the basis of self-evolving strategy, a new fuzzy rule base learning algorithm is presented to learn the rule base in MT2FLS: first, the rule base in T1FLS is learned, then it is extended to get the rule base in IT2FLS, finally we have the rule base in MT2FLS. For parameter learning, only the parameters of T1FLS in MT2FLS are learned by gradient descent and recursive least-squares with forgetting factor algorithms. The MT2FLS reduces complexity of designing an IT2FLS by reducing the computations in fuzzy rule base learning and the number of parameters needing to be learned without reducing the total number of parameters in MT2FLS. So the MT2FLS can obtain good performance with higher efficiency. Then the MT2FLS is applied to simulations on nonlinear plant modeling and chaotic signal prediction. Comparisons with other type-1 and type-2 fuzzy systems in both examples verify the performance of the MT2FLS.