Hybrid modeling of xanthan gum bioproduction in batch bioreactor

This work is focused on hybrid modeling of xanthan gum bioproduction process by Xanthomonas campestris pv. mangiferaeindicae. Experiments were carried out to evaluate the effects of stirred speed and superficial gas velocity on the kinetics of cell growth, lactose consumption and xanthan gum production in a batch bioreactor using cheese whey as substrate. A hybrid model was employed to simulate the bio-process making use of an artificial neural network (ANN) as a kinetic parameter estimator for the phenomenological model. The hybrid modeling of the process provided a satisfactory fitting quality of the experimental data, since this approach makes possible the incorporation of the effects of operational variables on model parameters. The applicability of the validated model was investigated, using the model as a process simulator to evaluate the effects of initial cell and lactose concentration in the xanthan gum production.