Application of fuzzy reasoning to control of glucose and ethanol concentrations in baker's yeast culture

Abstract

Fuzzy reasoning was applied to control both ethanol and glucose concentrations in fed-batch cultures of baker's yeast. This fuzzy controller consisted of three membership functions (concentrations of dissolved oxygen (DO), ethanol and glucose) and 18 production rules. Fuzzy inference was carried out by IF {A is a and B is b,...#x007D;, THEN {C is c} from the on-line measured concentrations of DO, ethanol and glucose. When “medium” concentrations of ethanol and glucose in fed-batch culture of baker's yeast were set at 2 g/l and 0.2 g/l, both ethanol and glucose concentrations were controlled at 2.67±0.35 g/l and 0.27±0.25 g/l, respectively, ethanol production was reduced from 26 g/l to 34 g/l, cell yield increased from 0.38 to 0.53 g dry cell/g consumed glucose and ethanol yield decreased from 0.50 to 0.14 g ethanol/g consumed glucose, respectively, as compared with those of the glucose only control at 0.2 g/l.