These days, in times of recession, we are forced by competitiveness and the
optimization of production to lower the costs of the temperature control in
pharmaceutical batch reactors and increase the quantity and quality of the
produced pharmaceutical product (active pharmaceutical substances).
Therefore, a control algorithm is needed which provides us rapid and precise
temperature control. This paper deals with the development of a control
algorithm, where two predictive functional controllers are connected in a
cascade for heating and cooling the content of the hybrid batch reactor. The
algorithm has to be designed to cope with the constraints and the mixed
discrete and continuous nature of the process of heating and cooling. The
main goal of the control law is to achieve rapid and exact tracking of the
reference temperature, good disturbance rejection and, in particular, a small
number of heating and cooling medium switchings. The simulation results of
the proposed algorithm give us much better performance compared to a
conventional cascade PI algorithm.