Fuzzy systems are able to treat uncertain and imprecise information; they m
ake use of knowledge in the form of linguistic rules. Their drawbacks are c
aused mainly by the difficulty of defining accurate membership functions an
d lack of a systematic procedure for the transformation of expert knowledge
into the rule base. Neural networks have the ability to learn but with som
e neural networks, knowledge representation and extraction are difficult. F
irst, we consider a rule-based fuzzy controller and a learning procedure ba
sed on the stochastic approximation method. The radial basis function neura
l network is then considered and it is shown that a modified form of this n
etwork is identical with the fuzzy controller, which may thus be considered
as a neuro-fuzzy controller. Numerical examples are presented to demonstra
te the validity of the approach and it is shown that such an adaptive neuro
-fuzzy system is successful in the control of a mobile robot. (C) 1999 Publ
ished by Elsevier Science B.V. All rights reserved.