Physics-constrained 3D convolutional neural networks for electrodynamics
APL Machine Learning · 01.06.2023 · 1 (2)
DOI: 10.1063/5.0132433 ↗Kurz gefasst
Fachartikel in der Zeitschrift APL Machine Learning, online erschienen am 14.04.2023. Verfasst von Alexander Scheinker und 1 weiteren Autor:innen. Die Arbeit wurde bisher 21-mal zitiert (Crossref) und nennt 72 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: High Energy Physics, Los Alamos National Laboratory.
Abstract (Originalsprache)
We present a physics-constrained neural network (PCNN) approach to solving Maxwell’s equations for the electromagnetic fields of intense relativistic charged particle beams. We create a 3D convolutional PCNN to map time-varying current and charge densities J(r, t) and ρ(r, t) to vector and scalar potentials A(r, t) and φ(r, t) from which we generate electromagnetic fields according to Maxwell’s equations: B = ∇ × A and E = −∇φ − ∂A/∂t. Our PCNNs satisfy hard constraints, such as ∇ · B = 0, by construction. Soft constraints push A and φ toward satisfying the Lorenz gauge.
Förderung
- High Energy Physics
- Los Alamos National Laboratory
Zitieren
APA
Scheinker, A., & Pokharel, R. (2023). Physics-constrained 3D convolutional neural networks for electrodynamics. APL Machine Learning, 1(2). https://doi.org/10.1063/5.0132433
Vancouver
Scheinker A, Pokharel R. Physics-constrained 3D convolutional neural networks for electrodynamics. APL Machine Learning. 2023;1(2). doi:10.1063/5.0132433
BibTeX
@article{Scheinker2023,
author = {Scheinker, Alexander and Pokharel, Reeju},
title = {Physics-constrained 3D convolutional neural networks for electrodynamics},
journal = {APL Machine Learning},
year = {2023},
volume = {1},
number = {2},
doi = {10.1063/5.0132433}
}
EUVIMED