Genetic Algorithm Tuned Fuzzy Neural Network for Non-linear System Control

      T.K. Radhakrishnan1) and   Madhubala T. Kadavarayar2)

1)Department of Chemical Engineering, National Institute of Technology, Tiruchirappalli 620 015, INDIA.  e-mail: radha@nitt.edu
2) Department of Physics,National Institute of Technology, Tiruchirappalli 620 015, INDIA.   e-mail: madhu_tk@rediffmail.com

Abstract

A fuzzy neural network controller is designed for a non-linear process. In this investigation, a real-time control of a conical liquid level system, which is a non-linear process, is studied. A conical tank is highly non-linear since the cross section of the tank varies with height. Here, the controller is tuned using genetic algorithms. Genetic algorithms are used to tune the membership functions of the input variables. The values of the parameters are constrained while running the genetic algorithms in order to get better results. The performance of the fuzzy neural network is compared to that of a well tuned PI controller.



To read the full article,
please log in:
If you haven't registered already,
you can do so for free:

© Copyright 2000-2004 Professor F.R. Hall & Dr I. Oraifige, University of Wolverhampton.