Application of deep learning methods combined with physical background in wide field of view imaging atmospheric Cherenkov telescopes  

作　　者：Ao-Yan Cheng Hao Cai Shi Chen Tian-Lu Chen Xiang Dong You-Liang Feng Qi Gao Quan-Bu Gou Yi-Qing Guo Hong-Bo Hu Ming-Ming Kang Hai-Jin Li Chen Liu Mao-Yuan Liu Wei Liu Fang-Sheng Min Chu-Cheng Pan Bing-Qiang Qiao Xiang-Li Qian Hui-Ying Sun Yu-Chang Sun Ao-Bo Wang Xu Wang Zhen Wang Guang-Guang Xin Yu-Hua Yao Qiang Yuan Yi Zhang 

机构地区：[1]School of Physics and Technology,Wuhan University,Wuhan 430072,China [2]School of Physics and Astronomy,Yunnan University,Kunming 650091,China [3]Ministry of Education,The Key Laboratory of Cosmic Rays(Tibet University),Lhasa 850000,China [4]Key Laboratory of Particle Astrophysics,Institute of High Energy Physics,Chinese Academy of Sciences,Beijing 100049,China [5]University of Chinese Academy of Sciences,19 A Yuquan Road,Shijingshan District,Beijing 100049,China [6]College of Physics,Sichuan University,Chengdu 610064,China [7]School of Intelligent Engineering,Shandong Management University,Jinan 250357,China [8]Tsung‑Dao Lee Institute,Shanghai Jiao Tong University,Shanghai 200240,China [9]Suzhou Aerospace Information Research Institute,Suzhou 215123,China [10]College of Physics,Chongqing University,No.55 Daxuecheng South Road,High‑tech District,Chongqing 401331,China [11]Key Laboratory of Dark Matter and Space Astronomy,Purple Mountain Observatory,Chinese Academy of Sciences,Nanjing 210008,China

出　　处：《Nuclear Science and Techniques》2024年第4期208-220,共13页核技术（英文）

摘　　要：The High Altitude Detection of Astronomical Radiation(HADAR)experiment,which was constructed in Tibet,China,combines the wide-angle advantages of traditional EAS array detectors with the high-sensitivity advantages of focused Cherenkov detectors.Its objective is to observe transient sources such as gamma-ray bursts and the counterparts of gravitational waves.This study aims to utilize the latest AI technology to enhance the sensitivity of HADAR experiments.Training datasets and models with distinctive creativity were constructed by incorporating the relevant physical theories for various applications.These models can determine the type,energy,and direction of the incident particles after careful design.We obtained a background identification accuracy of 98.6%,a relative energy reconstruction error of 10.0%,and an angular resolution of 0.22°in a test dataset at 10 TeV.These findings demonstrate the significant potential for enhancing the precision and dependability of detector data analysis in astrophysical research.By using deep learning techniques,the HADAR experiment’s observational sensitivity to the Crab Nebula has surpassed that of MAGIC and H.E.S.S.at energies below 0.5 TeV and remains competitive with conventional narrow-field Cherenkov telescopes at higher energies.In addition,our experiment offers a new approach for dealing with strongly connected,scattered data.

关 键 词：VHE gamma-ray astronomy HADAR Deep learning Convolutional neural networks 

分 类 号：P111[天文地球—天文学] TP18[自动化与计算机技术—控制理论与控制工程]