Extended Kalman Filtering Technique for the On-line Control of Candida Utilis Production

Abstract Extended Kalman Filter (EKF) is used for the on-line estimation of glucose, ethanol and biomass concentrations during the batch and fedbatch growth of Candida utilis on glucose. A physiological dynamic model describes the behavior of the yeast on glucose. The experimental system is presented. The performances of EKF are evaluated under both cultivation modes using classical observation variables: carbon dioxide production, oxygen and armnonia consumption. Although care has to be taken in the numerical procedures to obtain EKF algorithm convergence, successful on-line estimation of biomass and glucose concentrations are performed.