利用高速可微分模拟缩小机器学习与粒子加速器物理学之间的差距

IF 1.5 3区 物理与天体物理 Q3 PHYSICS, NUCLEAR
Jan Kaiser, Chenran Xu, Annika Eichler, Andrea Santamaria Garcia
{"title":"利用高速可微分模拟缩小机器学习与粒子加速器物理学之间的差距","authors":"Jan Kaiser, Chenran Xu, Annika Eichler, Andrea Santamaria Garcia","doi":"10.1103/physrevaccelbeams.27.054601","DOIUrl":null,"url":null,"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 <span>c</span>heetah, a <span>p</span>y<span>t</span>orch-based high-speed differentiable linear beam dynamics code. <span>c</span>heetah 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 <span>c</span>heetah as a user-friendly, readily extensible tool that integrates seamlessly with widely adopted machine learning tools. We showcase the utility of <span>c</span>heetah 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.","PeriodicalId":54297,"journal":{"name":"Physical Review Accelerators and Beams","volume":null,"pages":null},"PeriodicalIF":1.5000,"publicationDate":"2024-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Bridging the gap between machine learning and particle accelerator physics with high-speed, differentiable simulations\",\"authors\":\"Jan Kaiser, Chenran Xu, Annika Eichler, Andrea Santamaria Garcia\",\"doi\":\"10.1103/physrevaccelbeams.27.054601\",\"DOIUrl\":null,\"url\":null,\"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 <span>c</span>heetah, a <span>p</span>y<span>t</span>orch-based high-speed differentiable linear beam dynamics code. <span>c</span>heetah 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 <span>c</span>heetah as a user-friendly, readily extensible tool that integrates seamlessly with widely adopted machine learning tools. We showcase the utility of <span>c</span>heetah 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.\",\"PeriodicalId\":54297,\"journal\":{\"name\":\"Physical Review Accelerators and Beams\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":1.5000,\"publicationDate\":\"2024-05-28\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Physical Review Accelerators and Beams\",\"FirstCategoryId\":\"101\",\"ListUrlMain\":\"https://doi.org/10.1103/physrevaccelbeams.27.054601\",\"RegionNum\":3,\"RegionCategory\":\"物理与天体物理\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"PHYSICS, NUCLEAR\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Physical Review Accelerators and Beams","FirstCategoryId":"101","ListUrlMain":"https://doi.org/10.1103/physrevaccelbeams.27.054601","RegionNum":3,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"PHYSICS, NUCLEAR","Score":null,"Total":0}
引用次数: 0

摘要

机器学习已成为应对现代加速器物理挑战的强大解决方案。然而,有限的光束时间、模拟计算成本以及优化问题的高维性,都给生成训练最先进机器学习模型所需的数据带来了巨大挑战。在这项工作中,我们介绍了基于 pytorch 的高速可微分线性束流动力学代码 cheetah。cheetah 通过将计算时间减少多个数量级,实现了大型数据集的快速收集,并促进了加速器调谐和系统识别的高效梯度优化。这使得猎豹成为一款用户友好、易于扩展的工具,可与广泛采用的机器学习工具无缝集成。我们通过五个例子展示了猎豹的实用性,包括强化学习训练、基于梯度的光束线调整、基于梯度的系统识别、物理信息贝叶斯优化先验,以及空间电荷效应的模块化神经网络代理建模。使用这种高速可微分仿真代码将简化粒子加速器基于机器学习方法的开发,并快速将其集成到加速器设施的日常运行中。
本文章由计算机程序翻译,如有差异,请以英文原文为准。

Bridging the gap between machine learning and particle accelerator physics with high-speed, differentiable simulations

Bridging the gap between machine learning and particle accelerator physics with high-speed, differentiable simulations
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.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Physical Review Accelerators and Beams
Physical Review Accelerators and Beams Physics and Astronomy-Surfaces and Interfaces
CiteScore
3.90
自引率
23.50%
发文量
158
审稿时长
23 weeks
期刊介绍: Physical Review Special Topics - Accelerators and Beams (PRST-AB) is a peer-reviewed, purely electronic journal, distributed without charge to readers and funded by sponsors from national and international laboratories and other partners. The articles are published by the American Physical Society under the terms of the Creative Commons Attribution 3.0 License. It covers the full range of accelerator science and technology; subsystem and component technologies; beam dynamics; accelerator applications; and design, operation, and improvement of accelerators used in science and industry. This includes accelerators for high-energy and nuclear physics, synchrotron-radiation production, spallation neutron sources, medical therapy, and intense-beam applications.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信