PhD Positions in Scientific Machine Learning for Extreme Fluid Dynamics

In this project, three PhD candidates are sought to explore and develop cutting-edge scientific machine learning techniques that blends deep learning with physics-based techniques to control such extreme events in turbulent flows. Specifically, three capabilities are targeted:

(i) the identification of precursors and data-driven investigation of the mechanisms of extreme events using a blend of physical simulations and explainable AI techniques;

(ii) the forecasting of the turbulent flow before and throughout the extreme events using physics-constrained data-driven models able to self-correct;

(iii) the control of these flows to prevent extreme events, through a blend of model predictive control and deep learning techniques.