International Journal Bioautomation (Apr 2010)
An Application of Different Mixing Systems for Batch Cultivation of Saccharomyces cerevisiae. Part II: Multiple Objective Optimization and Model Predictive Control
Abstract
Multiple objective optimization of the initial conditions, maximal rotation speed and amplitude for a batch Saccharomyces cerevisiae cultivation using impulse and vibromixing systems is developed in this paper. The single objective function corresponds to the process productiveness and the residual glucose concentration. The multiple objective optimization problems are transformed to a single objective function with weight coefficients. A combined algorithm is applied for solving the single optimization. After this optimization the useful process productiveness increases and the residual glucose concentration at the end of the process decreases. The developed optimization and obtained results have shown that the impulse mixing systems have a better productiveness and better glucose assimilation. In addition, this system is easier for realization. The combined algorithm does not have a feedback and it does not guarantee robustness to process disturbances. For that purpose model predictive control for guarantee robustness to process disturbances is developed. The developed control algorithm - combined multiple objective optimization problem and model predictive control ensures maximal production at the end of the process and guarantees a feedback on disturbance as well as robustness to process disturbances.
