Optimizing Bioethanol Production via Extremum Seeking Control in a Continuous Stirred Tank Bioreactor

This chapter addresses bioethanol production in a continuous stirred tank bioreactor including inhibitory processes that might be present due to large substrate, biomass, and bioethanol concentrations. We aim to maximize the online productivity of the reactor despite the lack of knowledge of the mathematical model and external perturbations. To this end, as a reference, we determine the capability of the process to be optimized; thus, we use a mathematical model of the bioethanol production process. This mathematical model accounts for inhibition by modifying the reaction rates that are commonly considered. We fit the parameters of such a model with data from a typical bioethanol production scenario. In addition, we perform a sensitivity analysis to determine the key parameters regarding variations of the steady-state. By considering the dilution rate as the CSTR’s input and the productivity as output, we show the convexity of the input–output map. Such convexity shows that the bioethanol production process is suitable for optimizing its productivity using the dilution rate. A suitable control strategy to steer the CSTR to an optimal state is Extremum Seeking Control since it does not require the knowledge of the location of the optimal productivity nor the dynamical model. In order to show the applicability of the methodology, we use such an Extremum Seeking Control technique in two different ranges of species concentrations. The first one considers a large substrate concentration, typical in industrial applications, and the second one considers wastewater from tequila production as raw material for bioethanol production. Via numerical simulations, we show the performance of the extremum seeking control for both scenarios.