108 citations · 395 across the 21 of their papers we have counts for
23 papers
Heuristic Learning for Active Flow Control Using Coding Agents
Paul Garnier, Jonathan Viquerat, Elie Hachem
Active flow control involves nonlinear dynamics, partial observations, and computationally expensive simulations, making controller design particularly challenging. Deep reinforcem…
Graph Deep Learning for Intracranial Aneurysm Blood Flow Simulation and Risk Assessment
Paul Garnier, Pablo Jeken-Rico, Vincent Lannelongue +11
Intracranial aneurysms remain a major cause of neurological morbidity and mortality worldwide, where rupture risk is tightly coupled to local hemodynamics particularly wall shear s…
TRELLIS-Enhanced Surface Features for Comprehensive Intracranial Aneurysm Analysis
Clément Hervé, Paul Garnier, Jonathan Viquerat +1
Intracranial aneurysms pose a significant clinical risk yet are difficult to detect, delineate and model due to limited annotated 3D data. We propose a cross-domain feature-transfe…
Automated discovery of finite volume schemes using Graph Neural Networks
Paul Garnier, Jonathan Viquerat, Elie Hachem
Graph Neural Networks (GNNs) have deeply modified the landscape of numerical simulations by demonstrating strong capabilities in approximating solutions of physical systems. Howeve…
Training Transformers for Mesh-Based Simulations
Paul Garnier, Vincent Lannelongue, Jonathan Viquerat +1
Simulating physics using Graph Neural Networks (GNNs) is predominantly driven by message-passing architectures, which face challenges in scaling and efficiency, particularly in han…
Mixed-variable policy-based optimization
Jonathan Viquerat
The optimization of mixed-variable problems remains a significant challenge. We propose an extension of the policy-based optimization method that handles mixed-variables problems i…
Dragonfly: a modular deep reinforcement learning library
Jonathan Viquerat, Paul Garnier, Amirhossein Bateni +1
Dragonfly is a deep reinforcement learning library focused on modularity, in order to ease experimentation and developments. It relies on a json serialization that allows to swap b…
MeshMask: Physics-Based Simulations with Masked Graph Neural Networks
Paul Garnier, Vincent Lannelongue, Jonathan Viquerat +1
We introduce a novel masked pre-training technique for graph neural networks (GNNs) applied to computational fluid dynamics (CFD) problems. By randomly masking up to 40\% of input…
Multi-Grid Graph Neural Networks with Self-Attention for Computational Mechanics
Paul Garnier, Jonathan Viquerat, Elie Hachem
Advancement in finite element methods have become essential in various disciplines, and in particular for Computational Fluid Dynamics (CFD), driving research efforts for improved…
Beacon, a lightweight deep reinforcement learning benchmark library for flow control
Jonathan Viquerat, Philippe Meliga, Pablo Jeken +1
Recently, the increasing use of deep reinforcement learning for flow control problems has led to a new area of research, focused on the coupling and the adaptation of the existing…
Parallel bootstrap-based on-policy deep reinforcement learning for continuous flow control applications
J. Viquerat, E. Hachem
The coupling of deep reinforcement learning to numerical flow control problems has recently received a considerable attention, leading to groundbreaking results and opening new per…
Autoregressive Transformers for Data-Driven Spatio-Temporal Learning of Turbulent Flows
Aakash Patil, Jonathan Viquerat, Elie Hachem
A convolutional encoder-decoder-based transformer model is proposed for autoregressively training on spatio-temporal data of turbulent flows. The prediction of future fluid flow fi…
A review on deep reinforcement learning for fluid mechanics: an update
Jonathan Viquerat, Philippe Meliga, Elie Hachem
In the past couple of years, the interest of the fluid mechanics community for deep reinforcement learning (DRL) techniques has increased at fast pace, leading to a growing bibliog…
Graph neural networks for laminar flow prediction around random 2D shapes
Junfeng Chen, Elie Hachem, Jonathan Viquerat
In the recent years, the domain of fast flow field prediction has been vastly dominated by pixel-based convolutional neural networks. Yet, the recent advent of graph convolutional…
Robust deep learning for emulating turbulent viscosities
Aakash Patil, Jonathan Viquerat, George El Haber +1
From the simplest models to complex deep neural networks, modeling turbulence with machine learning techniques still offers multiple challenges. In this context, the present contri…
Policy-based optimization: single-step policy gradient method seen as an evolution strategy
Jonathan Viquerat, Régis Duvigneau, Philippe Meliga +2
This research reports on the recent development of a black-box optimization method based on single-step deep reinforcement learning (DRL), and on its conceptual proximity to evolut…
A twin-decoder structure for incompressible laminar flow reconstruction with uncertainty estimation around 2D obstacles
Junfeng Chen, Jonathan Viquerat, Frederic Heymes +1
Over the past few years, deep learning methods have proved to be of great interest for the computational fluid dynamics community, especially when used as surrogate models, either…
Deep reinforcement learning for the control of conjugate heat transfer with application to workpiece cooling
Elie Hachem, Hassan Ghraieb, Jonathan Viquerat +2
This research gauges the ability of deep reinforcement learning (DRL) techniques to assist the control of conjugate heat transfer systems governed by the coupled Navier--Stokes and…
U-net architectures for fast prediction of incompressible laminar flows
Junfeng Chen, Jonathan Viquerat, Elie Hachem
Machine learning is a popular tool that is being applied to many domains, from computer vision to natural language processing. It is not long ago that its use was extended to physi…
Exploiting locality and physical invariants to design effective Deep Reinforcement Learning control of the unstable falling liquid film
Vincent Belus, Jean Rabault, Jonathan Viquerat +3
Instabilities arise in a number of flow configurations. One such manifestation is the development of interfacial waves in multiphase flows, such as those observed in the falling li…
Direct shape optimization through deep reinforcement learning
Jonathan Viquerat, Jean Rabault, Alexander Kuhnle +3
Deep Reinforcement Learning (DRL) has recently spread into a range of domains within physics and engineering, with multiple remarkable achievements. Still, much remains to be explo…
A review on Deep Reinforcement Learning for Fluid Mechanics
Paul Garnier, Jonathan Viquerat, Jean Rabault +3
Deep reinforcement learning (DRL) has recently been adopted in a wide range of physics and engineering domains for its ability to solve decision-making problems that were previousl…
A supervised neural network for drag prediction of arbitrary 2D shapes in low Reynolds number flows
Jonathan Viquerat, Elie Hachem
Despite the significant breakthrough of neural networks in the last few years, their spreading in the field of computational fluid dynamics is very recent, and many applications re…