Gain-Scheduled Oxygen Concentration Control System for a Bio-Reactor

Abstract

This article presents a multi-variable control strategy for the regulation of Dissolved Oxygen Concentration in a stirred laboratory-scale bioreactor. The controller manipulates the airflow and agitation speed in the reactor. Departing from a mechanistic model that relates the manipulated variables and the oxygen transfer coefficient, linearized local models of the process are obtained. It is observed that there is high variability in the process behavior with the operating point. Then, a Gain-Scheduled controller is proposed. The strategy varies the agitation speed employing a PI control law and varies the air flow according to a switching law that detects saturations in the manipulated variables. The proposed strategy is evaluated in simulation and implemented in a real bioreactor. Results show that the strategy is able to regulate the oxygen concentration in a wide range of operating conditions.