Publications associées à Classification of tokamak plasma confinement states with convolutional recurrent neural networks

Deep learning approach for identification of H II regions during reionization in 21-cm observations - II. Foreground contamination

Michele Bianco, Jean-Paul Richard Kneib, Emma Elizabeth Tolley, Tianyue Chen

The upcoming Square Kilometre Array Observatory will produce images of neutral hydrogen distribution during the epoch of reionization by observing the corresponding 21-cm signal. However, the 21-cm signal will be subject to instrumental limitations such as ...

Oxford Univ Press2024

Global fluid simulations of plasma turbulence in stellarators

António João Caeiro Heitor Coelho

In order to cope with the decarbonization challenge faced by many countries, fusion is one of the few alternatives to fossil fuels for the production of electricity. Two devices invented in the middle of the previous century have emerged as the most promis ...

EPFL2024

Long plasma duration operation analyses with an international multi-machine (tokamaks and stellarators) database

Stefano Coda, Jeffrey Huang

Combined high-fusion performance and long-pulse operation is one of the key integration challenges for fusion energy development in magnetic devices. Addressing these challenges requires an integrated vision of physics and engineering aspects with the purp ...

IOP Publishing Ltd2024

First-principle based predictions of the effects of negative triangularity on DTT scenarios

Olivier Sauter, Gabriele Merlo, Alberto Mariani

Plasmas with negative triangularity (NT) shape have been recently shown to be able to achieve H-mode levels of confinement in L-mode, avoiding detrimental edge localised modes. Therefore, this plasma geometry is now studied as a possible viable option for ...

IOP Publishing Ltd2024

Generalization of Scaled Deep ResNets in the Mean-Field Regime

Volkan Cevher, Grigorios Chrysos, Fanghui Liu

Despite the widespread empirical success of ResNet, the generalization properties of deep ResNet are rarely explored beyond the lazy training regime. In this work, we investigate scaled ResNet in the limit of infinitely deep and wide neural networks, of wh ...

2024

Deep Learning Theory Through the Lens of Diagonal Linear Networks

Scott William Pesme

In this PhD manuscript, we explore optimisation phenomena which occur in complex neural networks through the lens of

2

-layer diagonal linear networks. This rudimentary architecture, which consists of a two layer feedforward linear network with a diagonal ...

EPFL2024

Error assessment of an adaptive finite elements-neural networks method for an elliptic parametric PDE

Alexandre Caboussat, Marco Picasso, Maude Girardin

We present a finite elements-neural network approach for the numerical approximation of parametric partial differential equations. The algorithm generates training data from finite element simulations, and uses a data -driven (supervised) feedforward neura ...

Elsevier Science Sa2024

Safe Deep Neural Networks

Kyle Michael Matoba

The capabilities of deep learning systems have advanced much faster than our ability to understand them. Whilst the gains from deep neural networks (DNNs) are significant, they are accompanied by a growing risk and gravity of a bad outcome. This is troubli ...

EPFL2024

Overview of fast particle experiments in the first MAST Upgrade experimental campaigns

Matteo Vallar

MAST-U is equipped with on-axis and off-axis neutral beam injectors (NBI), and these external sources of super-Alfv & eacute;nic deuterium fast-ions provide opportunities for studying a wide range of phenomena relevant to the physics of alpha-particles in ...

IOP Publishing Ltd2024

Experiments and gyrokinetic simulations of TCV plasmas with negative triangularity in view of DTT operations

Olivier Sauter, Stefano Coda, Alberto Mariani, Justin Richard Ball, Filippo Bagnato, Matteo Vallar

Negative triangularity (NT) scenarios in TCV have been compared to positive triangularity (PT) scenarios using the same plasma shapes foreseen for divertor tokamak test tokamak operations. The experiments provided a NT/PT L-mode pair and a PT H-mode with d ...

Iop Publishing Ltd2024

Towards practical reinforcement learning for tokamak magnetic control

Jonas Buchli, Federico Alberto Alfredo Felici, Cristian Galperti, Brendan Tracey

Reinforcement learning (RL) has shown promising results for real-time control systems, including the domain of plasma magnetic control. However, there are still significant drawbacks compared to traditional feedback control approaches for magnetic confinem ...

Elsevier Science Sa2024

Coupling a recurrent neural network to SPAD TCSPC systems for real-time fluorescence lifetime imaging

Claudio Bruschini, Edoardo Charbon, Andrei Ardelean, Paul Mos, Yang Lin

Fluorescence lifetime imaging (FLI) has been receiving increased attention in recent years as a powerful diagnostic technique in biological and medical research. However, existing FLI systems often suffer from a tradeoff between processing speed, accuracy, ...

Nature Portfolio2024

Inter-discharge optimization for fast, reliable access to ASDEX Upgrade advanced tokamak scenario

Olivier Sauter, Hartmut Zohm, Federico Alberto Alfredo Felici, Simon Van Mulders, Cassandre Ekta Contré

Rapid inter-discharge simulation and optimization using the RAPTOR code have allowed the development of a reliable and reproducible early heating strategy for an advanced tokamak (AT) scenario on ASDEX Upgrade. Solving for electron heat and current diffusi ...

IOP Publishing Ltd2024

Multi-machine benchmark of the self-consistent 1D scrape-off layer model DIV1D from stagnation point to target with SOLPS-ITER

Holger Reimerdes

This paper extends a 1D dynamic physics-based model of the scrape-off layer (SOL) plasma, DIV1D, to include the core SOL and possibly a second target. The extended model is benchmarked on 1D mapped SOLPS-ITER simulations to find input settings for DIV1D th ...

Iop Publishing Ltd2024

Random matrix methods for high-dimensional machine learning models

Antoine Philippe Michel Bodin

In the rapidly evolving landscape of machine learning research, neural networks stand out with their ever-expanding number of parameters and reliance on increasingly large datasets. The financial cost and computational resources required for the training p ...

EPFL2024

Re-design of EU DEMO with a low aspect ratio

Hartmut Zohm

The design point that had been chosen for EU DEMO in 2016 is reviewed here and a modification is proposed with a lower aspect ratio. Previously the same aspect ratio, A, was chosen for EU DEMO as in major tokamak experiments including ITER (A = 3.1), and, ...

2024

Error field detection and correction studies towards ITER operation

Olivier Sauter, Leonardo Pigatto, Bernhard Sieglin

In magnetic fusion devices, error field (EF) sources, spurious magnetic field perturbations, need to be identified and corrected for safe and stable (disruption-free) tokamak operation. Within Work Package Tokamak Exploitation RT04, a series of studies hav ...

IOP Publishing Ltd2024

Disruption avoidance and investigation of the H-Mode density limit in ASDEX Upgrade

Federico Alberto Alfredo Felici, Alessandro Pau, Bernhard Sieglin

In recent years a strong effort has been made to investigate disruption avoidance schemes in order to aid the development of integrated operational scenarios for ITER. Within the EUROfusion programme the disruptive H-mode density limit (HDL) has been studi ...

Iop Publishing Ltd2024

Nonlinear simulation of plasma turbulence using a gyrokinetic moment-based approach

Antoine Cyril David Hoffmann

Plasma turbulence plays a fundamental role in determining the performances of magnetic confinement fusion devices, such as tokamaks. Advances in computer science, combined with the development of efficient physical models, have significantly improved our u ...

EPFL2024

Divertor turbulence characterisation using Gas Puff Imaging in the TCV tokamak

Curdin Tobias Wüthrich

The performance of magnetic confinement fusion devices, such as tokamaks, is strongly correlated to the phenomena that occur in the boundary region of the plasma core that faces the wall of the device. The dominant cross-field transport mechanisms from the ...

EPFL2024