Neural fuzzy architecture for adaptive control

The authors address the tracking control of linear and nonlinear systems with unknown dynamics. A self-tuning neural fuzzy adaptive control architecture, based on M. Sugeno's model for fuzzy systems (1985) and on the use of feedforward neural networks is proposed. The control loop is described. Then, the adaptive neural version of the Sugeno model for fuzzy inference systems, inserted in this loop, is presented. The algorithm was simulated and tested for discrete-time linear and nonlinear systems.<
>