An adaptive fuzzy neural network for MIMO system modelapproximation in high-dimensional spaces

摘要

An adaptive fuzzy system implemented within the framework ofnneural network is proposed. The integration of the fuzzy system into anneural network enables the new fuzzy system to have learning andnadaptive capabilities. The proposed fuzzy neural network can locate itsnrules and optimize its membership functions by competitive learning,nKalman filter algorithm and extended Kalman filter algorithms. A keynfeature of the new architecture is that a high dimensional fuzzy systemncan be implemented with fewer number of rules than the Takagi-Sugenonfuzzy systems. A number of simulations are presented to demonstrate thenperformance of the proposed system including modeling nonlinearnfunction, operator's control of chemical plant, stock prices andnbioreactor (multioutput dynamical system)