Application of artificial neural network and fuzzy control for fed-batch cultivation of recombinant Saccharomyces cerevisiae

摘要: Abstract A recombinant Saccharomyces cerevisiae expressing β-galactosidase under the control of GAL10 promoter was constructed. The strain used as a model to study fuzzy with neural network, which served state estimators for fed-batch cultivation cells. To optimize expression effects medium enrichment and induction on cell growth were investigated. activity 2-fold higher in presence compared absence yeast extract basal medium. Furthermore, specific increased increasing galactose concentration up 30 g/ l . Two artificial networks (ANNs) developed estimate glucose concentrations using on-line measurements ethanol biomass concentrations, culture volume amount carbon source fed fermentor. improve productivity product yield two multi-variable controllers feed rates during production phases, respectively. Experimental data show that network estimators, 2.7-fold than case exponential feeding, 1.7-fold feeding feedback compensation concentration.