En savoir plus

The aim of this PhD project is to control the synthesis of aromas in wine by developing innovative real-time control strategies for the alcoholic fermentation process based on predictive mathematical models. These strategies will also have to take into account numerous constraints related to the control scheme, and the energy consumption of the fermentation, which must remain sufficiently low. In previous works, a dynamic model of wine alcoholic fermentation was developed, which represents the main kinetics, the synthesis of the main aromas, and the energy consumption of the process. Using this model, a simple MPC (Model Predictive Control) loop has been designed and tested on the real process, which constitutes a proof of concept of the approach. Depending on the background and interests of the PhD candidate, multiple research questions are open to explore:

• Biological modelling: development of mechanistic control-oriented models representing the production of multiple aroma compounds, and calibration using experimental data. Various system identification methods could be considered and proposed here. Additionally, model-reduction techniques can be tested, in particular based on a decomposition of slow/fast dynamics (considering the difference in speeds between the nitrogen and glucose metabolic pathways).
• Real-time control: design of novel control laws, taking into consideration aspects such as performance and robustness to model uncertainty. The range of methods can vary from nonlinear and possibly adaptive feedback control laws for systems with input constraints, to optimality-based control algorithms (like MPC loops). Observability issues are also of interest for online state estimation in real-time implementations.

The project will involve theoretical investigation of the mathematical problem, numerical simulations and experimental validation of control laws to be applied and tested on the real process located at the Pech Rouge Experimental Unit in Gruissan.