APPLICATION OF KALMAN FILTER ALGORITHMS IN A GMC CONTROL STRATEGY FOR FED-BATCH CULTIVATION PROCESS

Abstract

This paper addresses the study of a Generic Model Control (GMC) approach for regulation of the specific growth rate of for fed-batch cultivation process. This approach requires all the process states to be available on line. Since direct measurement of these states is not possible, a state estimator must be implemented. In this work, the estimator is a Kalman filter that uses the oxygen uptake rate (OUR) of the microorganisms as the measurable process output. Another practical issue constitutes the time delay present in the OUR measurements, a Smith predictor compensation is proposed to overcome this problem. Several simulations were carried out to test this scheme using an experimentally identified process model.