摘要
为消除或减弱γ射线对中子测量的影响,将脉冲堆积条件下的n脉冲信号和γ脉冲信号划分为单独n与γ脉冲事件、后沿堆积事件和前沿堆积事件共三种,根据复杂程度的不同,选取三种机器学习算法进行研究,最终采用基于支持向量机的n/γ甄别方法实现了波形甄别,并对最终的甄别效果进行分析,给出该方法的优势和下一步需要解决的问题。
In order to eliminate or reduce the influence ofγ- ray on neutron measurement, the n - pulse signal and γ- pulse signal under pile - up condition are divided into three types ： n and 7 pulse events, post - stac- king events and leading - edge stacking events. The three kinds of machine learning algorithms are selected to study. Finally, the n / γ discriminant method based on support vector machine is used to realize the pulse shape discrimination and the final result is analyzed. The advantage of this method and the next problems need to be solved are given.
出处
《核电子学与探测技术》
北大核心
2017年第1期8-11,共4页
Nuclear Electronics & Detection Technology
关键词
n/γ甄别
脉冲堆积
聚类分析
神经网络
支持向量机
n/γ discrimination
pulse pile - up
cluster analysis
neural networks
support vector machine
作者简介
王一鸣（1992-），男，湖北襄阳人，在读硕士研究生，攻读方向为辐射探测与防护研究。
