USE OF LEARNING AUTOMATA IN DISTRIBUTED FUZZY-LOGIC PROCESSOR TRAINING

Authors
IKONEN E NAJIM K
Citation
E. Ikonen et K. Najim, USE OF LEARNING AUTOMATA IN DISTRIBUTED FUZZY-LOGIC PROCESSOR TRAINING, IEE proceedings. Control theory and applications, 144(3), 1997, pp. 255-262
Citations number
20
Journal title
IEE proceedings. Control theory and applications → ACNP
ISSN journal
13502379
Volume
144
Issue
3
Year of publication
1997
Pages
255 - 262
Database
ISI
SICI code
1350-2379(1997)144:3<255:UOLAID>2.0.ZU;2-#
Abstract
A new algorithm for training the parameters of a distributed logic pro cessor is suggested, based on learning automata. A fuzzy inference sys tem is implemented on a multilayer perceptron platform. Various possib ilities for assigning learning automata are discussed and two assignme nt strategies are proposed. The behaviour of the resulting algorithm i s illustrated using flue-gas NOx emission data measured from an indust rial-size fluidised-bed combustor.