|
AUTOMATIC FUZZY RULE GENERATION IN A BIOTECHNOLOGICAL PROCESS
M. J. Fuente* G. I. Sainz Palmero** D. Pindado**
* Department of Systems Engineering and Control. Faculty of Science,
University of Valladolid. Prado de la Magdalena s/n,
47011 Valladolid, Spain. E-mail: maria@autom.uva.es
** ETSII. University of Valladolid. Paseo del Cauce s/n,
47011 Valladolid, Spain. E-mail:gresai@eis.uva.es.
This paper deals with the modeling of a biotechnological process from input-output data. A simple fuzzy-neural network is utilized to obtain a non-linear model of a wastewater plant. The fuzzy curves are utilized in order to know the significant input variables, the number of rules to model the system and the set of initial weights in the fuzzy neural network model. From the point of view of neural networks the model has weights to be trained from input output data and from the view point of fuzzy logic the model is a set of rules to explain the behaviour of the system.
Keywords: Fuzzy rules, neurofuzzy systems, modeling, wastewater plant
Session slot T-Fr-M04: Genetic algorithms and Rule generation/Area code 3e : Fuzzy and Neural Systems
 | 