Adaptive Control of S.cerevisiae Production

Abstract

The growth medium temperature of a bioreactor, in which S.cerevisiae was produced at aerobic condition, was controlled applying adaptive PID technique. Glucose was used as an energy and carbon source in the growth medium, and an exothermic reaction took place as a result of the glucose consumption of the microorganism under aerobic condition. Then the specific growth rate of the microorganism was inhibited by the increasing growth medium temperature. To maintain the growth medium temperature at its set point which ensures maximumyeast productivity and quality in the bioreactor, non-linear adaptive PID and linear adaptive PID control systems were employed numerically. The bioreactor in which S.cerevisiae yeast was produced under the aerobic condition was considered as a batch system with respect to mass balance. The heat balance between the cooling jacket and the growth medium in the bioreactor was considered as a continuous system with respect to energy. These approximationswere taken as a basis to develop the simulation work. In the simulation studies, models for the bioprocess were developed by using mass and energy balance equations and solved by utilizing the Runge–Kutta–Felhberg method. The control algorithm developed was added to the simulated model. For the adaptive PID control system, Non-linear Auto Regressive Moving Average with eXogenous (NARMAX) and Auto Regressive Moving Average with eXogenous (ARMAX) models were used, and relevant algorithms were developed on the basis of these black box models. The coefficients of the polynomials for both type of models were estimated from the relevant identification method. It was observed that non-linear adaptive PID control results were better than the results of linear adaptive PID and conventional PID control systems.