Abstract
In this paper a Distributed Adaptive Neuro-Fuzzy Architecture (DANFA) model with a second order Takagi-Sugeno inference mechanism is presented. The proposed approach is based on the simple idea to reduce the number of the fuzzy rules and the computational load, when modeling nonlinear systems. As a learning procedure for the designed structure a two-step gradient descent algorithm with a fixed learning rate is used. To demonstrate the potentials of the selected approach, simulation experiments with two benchmark chaotic time systems − Mackey-Glass and Rossler are studied. The results obtained show an accurate model performance with a minimal prediction error.