Model Predictive Control of Glucose Feeding for Fed-batch Candida utilis Biomass Production

In this work model predictive control (MPC) is used for control of glucose feeding rate of fed-batch culture for production of Candida utilis cell mass. A two layer hierarchical control system with the upper level for supervision and lower level for conventional control is designed which realizes automatic control, storage and refresh of kinetic model, data collection and on-line monitoring, data analysis and processing. The model stored in supervision computer is refreshed using genetic algorithm (GA) when the predictive error is over the limit. The results show that this method can control glucose concentration well and increase cell mass production effectively.