Control of nonlinear systems using non-stationary embedded recurrent fuzzy neural networks

Abstract

This study proposes a non-stationary embedded recurrent fuzzy neural network (NSRFNN) and its application on nonlinear system control. The NSRFNN preserves the ability of interval type-2 fuzzy systems with lower computational complexity. The NSRFNN has the concept of center variation of non-stationary fuzzy sets to enhance the performance of traditional membership functions. Finally, simulation results of nonlinear system control are shown to demonstrate the performance in computational effort of the proposed approach.