Feedback optimization of fed-batch baker's yeast fermentation
Abstract
A feedback optimization scheme for a fed-batch fermentation was developed. Saccharomyces cerevisiae, commonly known as baker's yeast, was chosen as a model system. As the first part of the proposed feedback optimization scheme, open-loop optimization studies have been carried out for the fed-batch baker's yeast culture. The simulation results demonstrated that the proposed feedback optimization scheme could handle the situations in which there are uncertainties in initial conditions and system parameters. The feedback optimization scheme could even overcome the changes in substrate feed concentration. A fully computerized fed-batch fermentation system has been built. It consisted of a 15 L fermentor equipped with a local controller; to which a microcomputer, on-line cell density measurement devices, and a feeding system were interfaced. A fermentor offgas line was connected to a mass spectrometer. The mass spectrometer was also interfaced to the microcomputer for control and data acquisition. The feedback optimization scheme has been implemented in low and high cell density fed-batch baker's yeast fermentations. The experimental results demonstrated the superiority of closed-loop implementation to the open-loop implementation for both low and high cell density runs. The performance indices in the closed-loop implementations were typically 10-20% greater than those in open-loop implementations. In the high cell density fermentations, experiments were carried out with an enriched oxygen source to overcome oxygen limitation, while measuring cell density on-line using a fiber optic device. A computational algorithm has been developed using the extended Kalman filter to estimate state variables and model parameters simultaneously. This algorithm was applied to a batch baker's yeast culture; based upon measurements of oxygen uptake rate, carbon dioxide evolution rate, and cell mass concentration. The Kalman filter yielded good estimates of the state variables, as well as kinetic model parameters, during exponential growth phases. A number of experimental studies have been carried out to investigate unequal uptake of $\alpha$-glucose and $\beta$-glucose by yeast. (Abstract shortened with permission of author.)
Degree
Ph.D.
Advisors
Lim, Purdue University.
Subject Area
Chemical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.