摘要
为建立TOPAZ-2型热离子反应堆电源中子物理快速计算方法,以对全范围仿真系统的堆芯计算提供新的辅助手段,采用神经网络方法,对堆芯的有效增殖系数keff和功率分布进行预测。通过MCNP得到不同转鼓角度组合下的高精度中子物理计算结果,将转鼓角度组合和对应的keff、功率分布结果分别作为输入和输出数据,建立代理模型。研究结果表明,代理模型得到的keff和功率分布都与MCNP结果符合较好,且计算时长大幅缩减。因此,研究建立的代理模型能够用于TOPAZ-2型热离子反应堆电源中子物理快速计算。
To establish a fast neutron physics calculation method for TOPAZ-2 thermionic reactor power supply,providing a new auxiliary means for core calculations of the full-scale simulator,this study employs neural network methods to predict the effective multiplication factor keff and core power distribution.The high-fidelity neutron physics calculation results under different drum angle combinations are obtained through MCNP,with the drum angle combinations and corresponding keff and power distribution results used as input and output data respectively,and a proxy model was established.The research results show that the keff and power distribution obtained by the surrogate model are in good agreement with the MCNP results,and reduce the length of the calculation at the same time.Therefore,the surrogate model established in this study can be used for fast neutron physics calculations in TOPAZ-2 thermionic reactor power supply.
作者
韩煦
姜强
杨宏伟
俞荣君
祁琳
Han Xu;Jiang Qiang;Yang Hongwei;Yu Rongjun;Qi Lin(China Institute of Atomic Energy,Beijing 102413,China)
出处
《电子技术应用》
2024年第S01期212-215,共4页
Application of Electronic Technique
基金
中国原子能科学研究院基金(PM2022082600000001)
关键词
堆芯
有效增值系数keff
功率分布
代理模型
reactor core
effective increment factor keff
power distribution
surrogate model
