Modeling and optimization IV: Investigation of reaction kinetics and kinetic constants using a program in which artificial neural network (ANN) was integrated


Bas D., DUDAK ŞEKER F. C. , Boyaci I. H.

JOURNAL OF FOOD ENGINEERING, cilt.79, ss.1152-1158, 2007 (SCI İndekslerine Giren Dergi) identifier identifier

  • Cilt numarası: 79 Konu: 4
  • Basım Tarihi: 2007
  • Doi Numarası: 10.1016/j.jfoodeng.2006.04.004
  • Dergi Adı: JOURNAL OF FOOD ENGINEERING
  • Sayfa Sayıları: ss.1152-1158

Özet

This work describes an application of artificial neural networks (ANNs) to determine kinetics of enzymatic reactions and to estimate kinetic constants. A model enzymatic reaction, the hydrolysis of maltose catalyzed by amyloglucosidase, was performed in a batch reactor and time courses were obtained. The artificial neural network was trained with the data of seven time courses and the other eight time courses were used for testing the network. The trained network was integrated in a script coded in MATLAB((R)), which is used for the selection of the proper kinetic model and its constants. The kinetics of the reaction was also investigated using the conventional method and the results were compared. The results of both methods imply that the uncompetitive inhibition kinetics was valid for the amyloglucosidase reaction and the kinetic constants (V-max, K-m and K-i) were 1.48 mu mol maltose/min/mg enzyme, 1.91 mM, and 71.42 mM for the model equation from developed program and 1.38 mu mol maltose/min/mg enzyme, 1.96 mM, and 94.34 mM for the model equation obtained from the conventional method. The usability of the model equation in a real engineering problem was also tested by a numerical solution of a differential equation obtained from the batch reactor. The time courses obtained from the developed program and conventional method were compared with the experimentally obtained time courses. The results indicate that the time courses obtained from the developed program fit more properly to the experimental data than that from conventional method. (c) 2006 Elsevier Ltd. All rights reserved.