% IMPORTANT: The following is UTF-8 encoded. This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.
@ARTICLE{Kaiser:600286,
author = {Kaiser, Jan and Xu, Chenran and Eichler, Annika and
Santamaria Garcia, Andrea},
title = {{B}ridging the gap between machine learning and particle
accelerator physics with high-speed, differentiable
simulations},
journal = {Physical review accelerators and beams},
volume = {27},
number = {5},
issn = {2469-9888},
address = {College Park, MD},
publisher = {American Physical Society},
reportid = {PUBDB-2023-07854, arXiv:2401.05815},
pages = {054601},
year = {2024},
note = {Phys. Rev. Accel. Beams 27 (2024) 054601. 16 pages, 9
figures, 3 tables},
abstract = {Machine learning has emerged as a powerful solution to the
modern challenges in accelerator physics. However, the
limited availability of beam time, the computational cost of
simulations, and the high dimensionality of optimization
problems pose significant challenges in generating the
required data for training state-of-the-art machine learning
models. In this work, we introduce cheetah, a pytorch-based
high-speed differentiable linear beam dynamics code. cheetah
enables the fast collection of large datasets by reducing
computation times by multiple orders of magnitude and
facilitates efficient gradient-based optimization for
accelerator tuning and system identification. This positions
cheetah as a user-friendly, readily extensible tool that
integrates seamlessly with widely adopted machine learning
tools. We showcase the utility of cheetah through five
examples, including reinforcement learning training,
gradient-based beamline tuning, gradient-based system
identification, physics-informed Bayesian optimization
priors, and modular neural network surrogate modeling of
space charge effects. The use of such a high-speed
differentiable simulation code will simplify the development
of machine learning-based methods for particle accelerators
and fast-track their integration into everyday operations of
accelerator facilities.},
keywords = {space charge (INSPIRE) / accelerator (INSPIRE) / machine
learning (INSPIRE) / optimization (INSPIRE) / reinforcement
learning (INSPIRE) / modular (INSPIRE) / Bayesian (INSPIRE)
/ neural network (INSPIRE)},
cin = {MSK},
ddc = {530},
cid = {I:(DE-H253)MSK-20120731},
pnm = {621 - Accelerator Research and Development (POF4-621) /
InternLabs-0011 - HIR3X - Helmholtz International Laboratory
on Reliability, Repetition, Results at the most advanced
X-ray Sources $(2020_InternLabs-0011)$ / ZT-I-PF-5-6 -
Autonomous Accelerator (AA) $(2020_ZT-I-PF-5-6)$},
pid = {G:(DE-HGF)POF4-621 / $G:(DE-HGF)2020_InternLabs-0011$ /
$G:(DE-HGF)2020_ZT-I-PF-5-6$},
experiment = {EXP:(DE-H253)ARES-20200101},
typ = {PUB:(DE-HGF)16},
eprint = {2401.05815},
howpublished = {arXiv:2401.05815},
archivePrefix = {arXiv},
SLACcitation = {$\%\%CITATION$ = $arXiv:2401.05815;\%\%$},
UT = {WOS:001237562300003},
doi = {10.1103/PhysRevAccelBeams.27.054601},
url = {https://bib-pubdb1.desy.de/record/600286},
}