Fuzzy control with fuzzy basis function neural network in magnetic bearing system

Huann-Keng Chiang Huann-Keng Chiang,Yong-Tang Jhou Yong-Tang Jhou,Chao-Ting Chu Chao-Ting Chu

doi:10.1109/isie.2012.6237199

Huann-Keng Chiang Huann-Keng Chiang, Yong-Tang Jhou Yong-Tang Jhou + Show 1 more

https://doi.org/10.1109/isie.2012.6237199

Copy DOI

Abstract

The paper propose the fuzzy control with recurrent fuzzy basis function neural network (RFBFNN) height control of magnetic bearing system. The magnetic bearing is a very unstable nonlinear system. Fuzzy control need not accurate mathematical model to have good output response for nonlinear systems. However, the fuzzy control cannot guarantee the stability and convergence. Neural networks also need not a precise mathematical model to describe the nonlinear systems. Neural network training is time consuming and unsuitable in real-time control. This paper uses the fuzzy control with the fuzzy basis function neural network to control the magnetic bearing system. The fuzzy basis function neural network use the gradient method for the system error and error derivative such that the output response has good response.

Full Text