Digital Science and Technology

Cooperative control of the driving trajectories for the energy efficiency of a fleet of vehicles

Vehicles that are communicating among each other and with a centralized infrastructure will be part of our future, in a gradual transition toward a fully connected system. In a connected vehicle world, deliberate exchange of intentions by vehicles and infrastructure reduces the need for guesstimating the surrounding traffic patterns and therefore enables better coordination. Automated vehicles can cooperate rather than compete for right of way in urban areas and highways, thus contributing to harmony in motion and improved mobility and efficiency of a group of vehicles.

Machine Learning for Efficient Numerical Simulations

Numerical simulation is an essential tool in order to model complex physical processes. Modern parallel codes implement complex algorithms that solve large systems of non-linear or partial differential equations. These computations produce a large amount of data that is usually discarded for future similar computations. The aim of this thesis is to leverage these data using machine learning in order to improve numerical simulation performance.

Large Eddy Simulation of spark ignition in aeronautical combustion chambers

During the development of an aeronautical gas turbine, one of the major concerns is to guaranty a correct reignition in case of extinction during the flight. Whereas manufacturers commonly use CFD (Computational Fluid Dynamics), and in particular LES (Large Eddy Simulation), to predict fuel consumption and pollutant emissions of gas turbines, these tools are rarely used to compute the reignition event due to its complexity and its high CPU cost. This explains why reignition is today essentially studied through experiments which remain extremely expansive and poorly instrumented.

A posteriori error analysis for complementarity problems

In this thesis, we are interested in the modeling of multiphase flow. This flow is described by a coupled system of nonlinear partial differential equations and nonlinear algebraic equations. The system of equations undergoes a wide variation of the data and presents phenomena of appearance and disappearance of phase. To manage these phenomena several formulations have been proposed: The Coats formulation is a dynamic management by variable switching where the retained equations and unknowns only relate to the presents phases.

Adaptive Downscaling Convolutional Neural Network

The use of tomography allows the volume reconstruction of porous materials in order to characterize them. It is well known that images from tomography, in particular from electronic tomography, are complex, noisy and very bulky. Today, the segmentation of these images is a tedious process, difficult to automate and extremely time consuming. This thesis aims to develop a suited and efficient method of segmentation to effectively determine the relevant descriptors of materials.

Design of Real-time Estimation Algorithms for Fault Detection and Load Mitigation Control at the Wind Farms Scale

In the field of wind energy, operators are now focusing on using existing wind farms more efficiently, reducing farm-level mechanical stress and reducing maintenance costs through improved fault detection. In this context, our central question will be "How to design an algorithm capable of optimally and robustly estimate the wake and 3D wind field in real-time at the wind farm scale?".

Mathematical and numerical study of coastal waves propagation at stratigraphic modeling scales

Describing natural processes that control our environment is a major challenge arising in numerous domains. In particular, describing long term erodability is mandatory to understand the impact of climate change on coastal landscapes. Such areas being among the most populated in the world, the social and economic issues are tremendous. Waves dynamic, mostly controlled by their frequency and amplitude, is the major engine of coastal erosion. Slope instabilities and coastal landslides depend on  their frequency and amplitude, as sediments transport from land to sea.

Analysis-Driven Design of Digital Multi-scale Microstructures of Materials

This thesis will focus on the creation of digital twins of microstructures of complex materials allowing in-fine retro-design of optimal materials in view of targeted usage properties. We will work on catalyst supports and building materials, focusing on the improvement of multi-physical properties taking into account transport properties and mechanical strength.