FEEDBACK IDENTIFICATION OF CONTINUOUS MICROBIAL-GROWTH SYSTEMS

The fundamental problem of dynamic modeling of continuous culture systems for process control and optimization is addressed. Forcing a system to bifurcation via feedback control is a very promising method for model discrimination and identification. Dynamic information is obtained by using feedback control to intentionally destabilize a steady state. Using this technique, the dynamic behavior of the chemostat as predicted by unstructured models, the Monod model with delay, and a structured model has been analyzed. The method exposes significant differences in the nonlinear dynamic structure of the various models and can be implemented to discriminate between various possible models for a continuous culture system.