摘要
为了消除硅微条探测器的电荷响应对入射位置和入射方向的依赖,提高电荷分辨能力,使用OPTICS、高斯混合与主成分分析三种机器学习算法协同完成了电荷重建,并与经典方法进行对比。结果表明,相较于经典方法,基于机器学习重建的电荷谱峰谷比更大、电荷分辨更小,重建效果更好。
Three machine learning algorithms,OPTICS,Gaussian mixture and principal component analysis,are introduced to improve the charge resolution of silicon micro-strip detectors by correcting the charge response dependence on the incident position and direction.Compared with the classical method,the charge spectrum using the machine learning algorithms has a larger peak to valley ratio and better charge resolution.
作者
闵江洪
乔锐
于龙昆
彭文溪
龚轲
郭东亚
崔兴柱
刘雅清
MIN Jiang-hong;QIAO Rui;YU Long-kun;PENG Wen-xi;GONG Ke;GUO Dong-ya;CUI Xing-zhu;LIU Ya-qing(School of Information Engineering,Nanchang University,Nanchang 330031,China;Institute of High Energy Physics,Chinese Academy of Sciences,Beijing 100049,China;School of Advanced Manufacturing Engineering,Nanchang University,Nanchang 330031,China)
出处
《核电子学与探测技术》
CAS
北大核心
2023年第2期318-324,共7页
Nuclear Electronics & Detection Technology
基金
国家自然科学基金天文联合基金项目(U1738133)资助。
关键词
硅微条探测器
电荷重建
机器学习
silicon micro-strip detector
charge reconstruction
machine learning
作者简介
闵江洪(1998-),女,湖南益阳人,在读硕士,攻读方向为硅微条探测器及应用;通讯作者:乔锐,男,副研究员,E-mail:qiaorui@ihep.ac.cn。
