YEAST CONCENTRATION ESTIMATION AND PREDICTION WITH STATIC AND DYNAMICNEURAL-NETWORK MODELS IN BATCH CULTURES

The second fermentation is one of the most important steps in Champagn e production. For this purpose, yeasts are grown on a wine based mediu m to adapt their metabolism to ethanol. Several models built with vari ous static and dynamic neural network configurations were investigated . The main objective was to achieve real-time estimation and predictio n of yeast concentration during growth. The model selected, based on r ecurrent neural networks, was first order with respect to the yeast co ncentration and to the volume of CO2 released. Temperature and pH were included as model parameters as well. Yeast concentration during grow th could thus be estimated with an error lower than 3% (+/- 1.7 x 10(6 ) yeasts/ml). From the measurement of initial yeast population and tem perature, it was possible to predict the final yeast concentration (af ter 21 hours of growth) from the beginning of the growth, with about /- 3 x 10(6) yeasts/ml accuracy. So a predictive control strategy of t his process could be investigated.