Model Discrimination and Parameter Identification by an Experimental Design Strategy

Abstract

An experimental design concept for the discrimination of unstructured models and for the identification of model parameters is presented.Using the anaerobic bacterium Zymomonas mobilis the quality of this design is shown by simulations. In case of well-known kinetic modeling approaches a D-optimal design strategy for steady-state fermentations is presented. This sequential experimental design was applied for the identification of macrokinetics of the methylotrophic yeast Candida boidinii. Nutristatic process control is used for the realization of precalculated operating points.