Production optimization with operating constraints for a fed-batch reactor with DMC predictive control

This work presents a production optimization approach with operating constraints applied to a fed-batch penicillin process with a DMC predictive controller. A deterministic and non-structured mathematical model of the fed-batch penicillin production process solved by a fourth order variable step Runge-Kutta-Gill algorithm type simulated the system. The production process optimization has been performed through the manipulation of the initial cell and the substrate fed-batch concentrations besides the fed-batch flow rate strategy. In order to carry out this procedure it was necessary to compute a production index that considers product amount and process operating constraints such as cell, substrate and dissolved oxygen concentrations besides the bioreactor final volume through the penalty functions concept. The optimization algorithm was based on the modified simplex method. The DMC control algorithm was implemented in order to stabilize the dissolved oxygen concentration through the agitation speed manipulation. This controller was found to be appropriate to work coupled with a optimization procedure due to the fact that it gives a smooth behavior to this critical constraint even with all disturbs during the convergence. (C) 1999 Elsevier Science Ltd. All rights reserved.