Observer (state estimator) design for a winemaking fermentation process

In the global framework of the DigitWine project, the proposed PhD will focus on the design of an observer (state estimator) of the wine fermentation process. Since wine fermentation is subject to strong variability of the biological material (grapes, yeasts) and the existing dynamic models, however accurate, are still imperfect, available on-line measurements will be used to update the knowledge of the current process state in real time. Based on a literature survey, the PhD candidate will explore several options, ranging from classical observers such as Luenberger observer and Kalman filter to modern tools based on artificial intelligence [1,2] or fuzzy logic [3]. The advantages and limitations of various options will be compared and a number of promising strategies will be explored more in-depth. The work will start with existing dynamical models of the process and available experimental data and will be progressively enriched and updated as the partners of the DigitWine project will develop new models and acquire additional experimental data. The designed observer will be integrated in the digital twin developed for the wine fermentation process. [1] Alexander R et al. (2020). Challenges and opportunities on nonlinear state estimation of chemical and biochemical processes, Process 8(11) 1462. [2] Feng S et al. (2023). A review: state estimation based on hybrid models of Kalman filter and neural network, Systems Science & Control Engineering, 11(1) 2173682. [3] Marquez-Vera et al. (2018). Stable fuzzy control and observer via LMIs in a fermentation process, J of Computational Sci, 27 192-198.