COMPUTER CONTROL AND ADAPTIVE ON-LINE OPTIMIZATION OF A CONTINUOUS BIOREACTOR

MICHAEL JAMES ROLF, Purdue University

Abstract

Methods for the control and optimization of a continuous bioreactor were developed. Specifically, the maximization of cellular productivity in a continuous bakers' yeast fermentation was considered. Four problems were dealt with: bioreactor start-up; finding the point of optimum cellular productivity; tracking the optimum as changes occur in the process; and bioreactor stability. An adaptive on-line optimization method was developed which employed dynamic model identification. This method is attractive in the application to continuous bioreactors for the following reasons: very little a priori information is needed; it is easy to implement; it is faster than conventional methods; it adapts to changes in the process; and it is very stable. Computer simulations of the optimization of a continuous bakers' yeast fermentation were performed. These showed feasibility of the method and also yielded approximate values of tuning factors necessary for experimental implementation. A versatile computerized fermentation system was developed to facilitate experimental investigations and to implement control and optimization schemes. The computerized fermentation system consists of two parts: a real-time computer control system; and a highly instrumented bioreactor with complete feedback capabilities. Experiments were conducted which tested and compared various startup procedures for a continuous bakers' yeast fermentation. Oscillatory phenomena in continuous bakers' yeast fermentations was also investigated, particularly with respect to the dissolved oxygen control loop. Finally, the adaptive on-line optimization method was implemented experimentally. Results show that the method optimizes quickly, adapts to changes in the process, and is stable even with the occurrence of operational difficulties.

Degree

Ph.D.

Subject Area

Chemical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS