General fuzzy models for dynamical systems

Fuzzy logic systems can approximate any nonlinear function and they can also be successfully applied to system modeling. This paper describes an efficient fuzzy identification method which can be applied to nonlinear dynamical systems. The proposed method is constituted of three steps: in the first step, an ordinary fuzzy identification technique is applied, steps 2 and 3 are introduced in order to improve the model obtained in step 1, and then a model with higher resolution is obtained. To demonstrate the effectiveness of the proposed method, several examples are presented.