摘要
针对相同几何形状、等质量、不同丰度的金属铀部件,采用^(252)Cf源驱动噪声分析法来获得中子时间关联计数。通过对中子时间关联计数的分析、处理,确定特征参数。采用BP神经网络方法通过一定数量的训练后,对未知丰度的金属铀部件进行判定。结果显示,采用BP神经网络方法可以对金属铀部件的丰度进行有效地判定。该方法可应用于金属铀部件身份认证工作。
To identify metal uranium components with same geometry and mass but different enrichments, 252Cf source noise analysis was used to gain the neutron time correlation coincidence measurement. Based on the analysis and processing of neutron time correlation coincidence measurement, the characteristic parameters are determined. Through the method of BP neural network, we determined the 235U enrichments of same geometry and mass but different enrichments uranium components. Result shows that by the method of BP neural network, we can diagnose the 235U enrichments effectively.
作者
任立学
刘知贵
周之入
Ren Lixue;Liu Zhigui;Zhou Zhiru(Institute of Electronic Engineering,China Academy of Engineering Physics,Mianyang,Sichuan 621900,China;School of National Defense Science and Technology,Southwest University of Science and Technology,Mianyang,Sichuan,621010,China;Nuclear Power Institute of China,Chengdu,610213,China)
出处
《核动力工程》
EI
CAS
CSCD
北大核心
2019年第1期48-50,共3页
Nuclear Power Engineering
基金
国家自然科学基金NSAF(No 11176031)
关键词
中子时间关联计数
BP神经网络
丰度
Neutron time correlation coincidence
BP neural network
Enrichment
