摘要
利用蒙特卡罗(Monte Carlo)模拟产生的数据研究了由ARGO探测阵列所观测的,由能量为100GeV—10TeV,天顶角为0°—45°的原初γ射线和强子(质子和原子核)所引起的簇射的横向分布结构,得到了平均横向分布宽度、最小树长度等可以描述两种簇射空间分布差异的特征量.研究了用这些特征量作为输入单元的人工神经网络分析方法进行原初γ和强子分辨的能力,结果表明,利用该方法可有效地区分γ和强子簇射.
The lateral distributions, as measured by ARGO array, of the extensive air showers induced by y and hadrons with energy range from 100 GeV to 10 TeV, zenith angle from 0°to 45°, were studied using Monte Carlo simulated data. Several parameters such as average lateral distribution, minimum tree length etc., which could be used to distinguish the lateral distributions between the showers induced by γ and hadrons, were obtained. These parameters were used as the input for an artificial neural network, which was then trained to study the γ/hadron discrimination power. The result indicated that using this method could effectively separate.
出处
《高能物理与核物理》
EI
CSCD
北大核心
2005年第5期485-490,共6页
High Energy Physics and Nuclear Physics
基金
国家自然科学基金(10475051
10120130794)
山东省自然科学基金(Y2002A07)资助~~
关键词
蒙特卡罗模拟
人工神经网络
ARGO实验
最小树长度
强子
Γ射线
ARGO experiment, artificial neural network,γ-ray identification,shower lateral distribution, minimum tree length