Neural control of an imperfectly mixed fed-batch bioreactor for recombinant β-galactosidase

Patnaik, P R (1999) Neural control of an imperfectly mixed fed-batch bioreactor for recombinant β-galactosidase. Biochemical Engineering Journal, 3 (2). pp. 113-120. ISSN 1369703X

PDF patnaik1999.6.pdf - Published Version Restricted to Registered users only Download (167Kb) | Request a copy

Abstract

The production of β-galactosidase in fed-batch fermentation by a recombinant strain of Escherichia coli was studied for on-line optimization by PID control and by an artificial neural network (ANN). The process was represented by an Elman ANN and the controller by a feed-forward network. Based on earlier work, the fermentation was carried out below a threshold temperature in an imperfectly mixed bioreactor. To mimic an industrial situation, data generated by a deterministic model were corrupted by adding 10% Gaussian noise. For dynamic optimal performance, the controller adjusted the mixing characteristics for each sampling interval on the basis of data for the previous interval. Application of neural control increased the maximum attainable concentration of β-galactosidase by 23.8% per gram of cell mass and by 51.1% per unit volume of the broth, compared with the uncontrolled fermentation; corresponding increases with adaptive PID control were 8.1% and 17.7%, respectively.

Item Type:	Article
Additional Information:	Copyright of this article belongs to Elsevier Science.
Uncontrolled Keywords:	Recombinant fermentation; β-Galactosidase; Imperfect mixing; Neural control; PID control
Subjects:	Q Science > QR Microbiology
Depositing User:	Dr. K.P.S.Sengar
Date Deposited:	31 Jan 2012 12:30
Last Modified:	12 Jan 2015 04:22
URI:	http://crdd.osdd.net/open/id/eprint/829

Actions (login required)

View Item