15th Triennial World Congress of the International Federation of Automatic Control
  Barcelona, 21–26 July 2002 
FUZZY IDENTIFICATION OF BIOPROCESSES APPLYING TSK-TYPE MODELS WITH CONSEQUENT PARAMETER ESTIMATION THROUGH ORTHOGONAL ESTIMATOR
Rosimeire Aparecida Jerônimo*,1, Claudio Garcia*,1,
Oscar A. Z. Sotomay or**
* Department of Telecommunications and Control Engineering,
Laboratory of Automation and Control (LAC)
** Department of Chemical Engineering, Laboratory of Simulation
and Process Control (LSCP)
Polytechnic School of the University of São Paulo
Av. Prof. Luciano Gualberto, Trav. 3, n. 158, Cidade
Universitária, CEP 05508-900, São Paulo - SP, Brazil
This paper presents fuzzy identification of two bioprocesses employing TSK-type models. A “Modified Gram-Schmidt” (MGS) orthogonal estimator is used to estimate the consequent parameters. This approach is then applied to identify two distinct cases involving dissolved oxygen concentration: one related to a bioreactor and the other one related to an activated sludge process. The obtained models are then cross-validated.
Keywords: System identification, Bioprocesses, Nonlinear models, Fuzzy modeling, TSK-type models, Modified Gram-Schmidt algorithm, Parameter estimation
1Corresponding authors. Tel: +55-11-3818-5648; Fax: +55-11-3818-5718 E-mail address: rosi@lac.usp.br (R.A. Jerônimo), (C. Garcia)

E-mail: clgarcia@lac.usp.br
Session slot T-Fr-A04: Industrial Applications of Fuzzy and Neural Systems/Area code 3e : Fuzzy and Neural Systems