An adaptive fuzzy-neural controller for multivariable system

Abstract

This paper presents an adaptive fuzzy-neural controller for multivariable system which incorporates the advantage of fuzzy logic and neural network. Inverted pendulum is well known as a multivariable and nonlinear system. And it is very difficult to design and realize a single stage fuzzy controller for the problem of multivariable system as inverted pendulum. After the description of the research background of quadruple inverted pendulum system, a fuzzy-neural controller is proposed which is based on three dimension fuzzy composed variable and the parameters are optimized by backpropagation algorithm in neural network. And the parameters could be adjusted online further. The proposed method gives a solution to design and realize high dimension fuzzy controller in multivariable system. Simulation shows its validity