计算物理 ›› 1999, Vol. 16 ›› Issue (6): 568-572.

• 论文 • 上一篇    下一篇

用神经网络分辨“膝”区原初质子成分

冯存峰, 孔繁敏, 张学尧, 何瑁, 戴志强, 张乃健   

  1. 山东大学物理系高能物理研究室, 济南 250100
  • 收稿日期:1998-09-29 修回日期:1999-05-17 出版日期:1999-11-25 发布日期:1999-11-25
  • 作者简介:冯存峰,男,30,讲师,博士
  • 基金资助:
    山东省自然科学基金和国家自然科学基金资助项目

Identification of primary proton component at the knee with artificial neural networks

Feng Cunfeng, Kong Fanmin, Zhang Xueyao, He Mao, Dai Zhiqiang, Zhang Naijian   

  1. High energy physics group, physics department, Shandong University, Jinan 250100
  • Received:1998-09-29 Revised:1999-05-17 Online:1999-11-25 Published:1999-11-25

摘要: 利用不同的相互作用模型,对羊八井空气簇射阵列和乳胶室联合实验进行了Monte Carlo模拟,并利用模拟数据研究了利用人工神经网络方法进行原初宇宙线成分分辨的可行性。分析表明,利用该方法可很好地挑选出由原初质子引起的簇射,但所得结果与作用模型有关,并对相应的系统误差进行了估计。

关键词: Monte Carlo模拟, 神经网络, 原初质子分辨

Abstract: A Monte Carlo simulation for the hybrid experiment of air shower array and emulsion chambers at Yangbajing is done with different interaction models.The feasibility of distinguishing primary cosmic ray components by use of the method of artificial neural netwo rks is studied with the simulation data.T he analysis indicates that the showers induced by primary protons can be efficiently selected by using this method, but the results obtained appear to be dependent on interaction model.The corresponding systematic error is also estimated.

Key words: Monte Carlo simulation, artificial neural networks, identification of primary protons

中图分类号: