The seminar will present the results of a project aimed at examining the feasibility of using a modern control approach to regulating diacetyl concentration in order to minimise the variance in batch length, using temperature as the manipulated variable. Analytical models of sugar, yeast and alcohol concentrations, and neural network models of diacetyl concentration have been developed and validated against commercial beer plant data. These models have been used together in a simulated model-based predictive controlled system. The results have been encouraging and plant trials are planned.
Back to Control Seminars Page.