基于Python的PFA光电子径迹重建  

PFA Photoelectron Track Reconstruction Based on Python

作　　者：陈晓 杜园园[2] CHEN Xiao;DU Yuanyuan(College of Science,Tibet University,Lhasa Tibet 850000;Key Laboratory of Particle Astrophysics,Institute of High Energy Physics,Chinese Academy of Science,Beijing 100049)

机构地区：[1]西藏大学理学院,西藏拉萨850000 [2]中国科学院高能物理研究所粒子天体中心,北京100049

出　　处：《软件》2022年第2期103-107,共5页Software

摘　　要：探测天文X射线的偏振有很重要的物理意义,气体像素探测器(GPD)是探测软X射线(2-10keV)偏振的主要工具。光电效应中,光电子的发射方向与入射X射线的偏振密切相关,所以可以通过分析光电子的出射方向来探测X射线的偏振情况。本文介绍了两种重建光电子发射方向的方法,分别是解析惯性主轴方向和机器学习预测光电子发射方向。这两种重建方法通过统计光电子发射方向均能很好的判断出入射X射线的偏振方向,而且随着机器学习框架和深度学习算法的迅速发展,机器学习预测光电子发射方向的方法将会有更好的发展潜力和更高的预测准确性上限。It is of great physical significance to detect the polarization of astronomical X-rays.Gas pixel detector(GPD)is the main tool to detect the polarization of soft X-rays(2-10keV).In photoelectric effect,the emission direction of photoelectrons is closely related to the polarization of incident X-rays,so the polarization of X-rays can be detected by analyzing the emission direction of photoelectrons.In this paper,two methods of reconstructing photoelectron emission direction are introduced,namely,analyzing the inertial principal axis direction and machine learning to predict the photoelectron emission direction.Both of the two reconstruction methods can determine the polarization direction of incident X-rays by counting the photoelectron emission direction.These two reconstruction methods can well judge the polarization direction of incoming and outgoing X-rays by counting the photoelectron emission direction.With the rapid development of machine learning framework and deep learning algorithm,the method of predicting photoelectron emission direction by machine learning will have better development potential and higher upper limit of prediction accuracy.

关 键 词：X射线偏振 气体像素探测器 重建方向 机器学习 灰度图方向 

分 类 号：V524[航空宇航科学与技术—人机与环境工程]