未来电子-离子对撞机的AI优化探测器设计：双 radiator RICH案例
仪器与探测器
2020-06-11 v2 机器学习
高能物理 - 实验
摘要
先进的探测器研发需要在探测器设计优化过程中进行计算密集且细致的模拟。我们提出一种基于贝叶斯优化和机器学习、并编码探测器需求的通用优化流程。作为一个案例研究，我们聚焦于未来电子-离子对撞机（EIC）粒子鉴别系统研发中的双 radiator Ring Imaging Cherenkov（dRICH）探测器设计。EIC是一项美国主导的核物理前沿加速器项目，旨在进一步探索海夸克和胶子尺度下核物质的结构与相互作用。我们展示在现有模型假设下，通过我们自动化且高度并行的框架获得的探测器设计优于基线dRICH设计。只要具备真实模拟，我们的方法可应用于任何探测器研发。
引用
@article{arxiv.1911.05797,
title = {AI-optimized detector design for the future Electron-Ion Collider: the dual-radiator RICH case},
author = {E. Cisbani and A. Del Dotto and C. Fanelli and M. Williams and M. Alfred and F. Barbosa and L. Barion and V. Berdnikov and W. Brooks and T. Cao and M. Contalbrigo and S. Danagoulian and A. Datta and M. Demarteau and A. Denisov and M. Diefenthaler and A. Durum and D. Fields and Y. Furletova and C. Gleason and M. Grosse-Perdekamp and M. Hattawy and X. He and H. van Hecke and D. Higinbotham and T. Horn and C. Hyde and Y. Ilieva and G. Kalicy and A. Kebede and B. Kim and M. Liu and J. McKisson and R. Mendez and P. Nadel-Turonski and I. Pegg and D. Romanov and M. Sarsour and C. L. da Silva and J. Stevens and X. Sun and S. Syed and R. Towell and J. Xie and Z. W. Zhao and B. Zihlmann and C. Zorn},
journal= {arXiv preprint arXiv:1911.05797},
year = {2020}
}
备注
22 pages, 11 figures