Application of knowledge information engineering for sake mashing process

Simulation and control using a mathematical model are often difficult in the sake mashing process, because this process is a complicated process which involves many microorganisms and enzymes. Recently, knowledge information processings, such as fuzzy logic, artificial neural network (ANN), and genetic algorithm (GA), have been developed. These information processings have been applied to control sake mashing process. The fuzzy logic and fuzzy neural network with the extraction of toji's knowledge and experience about the temperature control of the sake mashing process were applied to the temperature control of experimental mashing. Time course data were similar to those from a conventional control based on the decision of the toji. ANN was applied to estimate enzyme activities in koji from the temperature and humidity orbits of koji making process. The suitable courses of temperature and humidity for koji production with the desired values of enzyme activities were determined by applying these models and GA.