Monitoring Ethanol Fermentation in Real Time by a Robust State Observer for Uncertainties.

In this article, the problem of real‐time estimation of the fermentative ethanol process is tackled. The considered observer is a model‐based technique that is robust regarding the model parameter uncertainties and inline noisy measurements. An unstructured kinetic model was used to describe the production of ethanol in a batch bioreactor for Saccharomyces cerevisiae. The biomass concentration was selected as the measured bioreactor´s output via an inline device, where the estimate variables were the substrate and ethanol concentrations. An experimental prototype was constructed to demonstrate the observer's real‐time performance. The experimental results show that the robust, smooth sliding mode observer performs better than the standard proportional sliding mode observer. [ABSTRACT FROM AUTHOR]