A fuzzy neural hybrid system
Rajab Challoo, Dominic A. Clark, Robert A. McLauchlan, S.I. Omar · 2002
This paper presents an architecture for a fuzzy neural hybrid system (FNHS). This architecture merges fuzzy logic and neural networks by replacing the rule-base of a typical fuzzy logic controller with a backpropagation neural network. A method is developed for calculating the errors needed to train the neural network. Simulation results are presented which show that the FNHS performs as well as or better than a conventional neural network. Other advantages of using this system are also presented.>