AN ADAPTIVE OPTIMAL-CONTROL SCHEME BASED ON HYBRID NEURAL MODELING

A hybrid neural modelling procedure which enables the implementation o f an adaptive control scheme for the optimization of fed-batch ferment ations is presented. Simulations for the processes of cell mass produc tion and ethanol fermentation by Sacharomyces cerevisae show that, in the presence of modelling errors, the adaptive control leads to nearly optimal results, while open-loop control leads to bad results. Experi mental studies show that, for the process of ethanol fermentation by Z ymomonas mobilis, a hybrid neural model can be-developed with relative ly few experimental data and the use of an approximate mathematical mo del. (C) 1998 Published by Elsevier Science Ltd. All rights reserved.