Application of High Gain and Sliding Mode Observers for Estimating Key Variables in Bioprocesses

Abstract

The monitoring and control of bioreactors is essential for any bioprocess. Therefore, measuring some essential biochemical key variables is crucial, such as the reaction rates, whose structure is usually barely known. Although many high gain and sliding mode observers-based estimators were designed to estimate these variables, these strategies are not robust against external disturbances. This work extends the high gain and sliding mode techniques to simultaneously estimate the reaction rates and an external disturbance under a persistent excitation condition that is easy to verify. Furthermore, the proposed algorithms are exponentially and finite-time stable, respectively. Finally, both observers are tested in a continuous fermentation process simulation.