反应堆Monte Carlo临界计算加速收敛方法综述  

Review of convergence acceleration methods in Monte Carlo criticality calculations for reactor analysis

作　　者：潘流俊[1] 王瑞宏[1] 江松[1] 许海燕[1] 上官丹骅[1] 姬志成[1] 

机构地区：[1]北京应用物理与计算数学研究所,北京100094

出　　处：《中国科学：信息科学》2016年第10期1489-1509,共21页Scientia Sinica(Informationis)

基　　金：国家重点基础研究发展计划(973计划)(批准号:2011CB309705)资助项目

摘　　要：反应堆临界计算的核心问题是计算系统的裂变源分布,有了裂变源分布以后,就能够很方便地得到有效增殖因子、系统的功率分布和反应率等物理量.在反应堆计算和分析中,Monte Carlo方法在处理复杂几何和中子参数等方面具有独特的优势.随着计算机软硬件能力的不断发展,Monte Carlo临界计算已经逐步应用于实际反应堆的全堆芯模拟,同时有关这方面的研究已经发展成一个热点.Monte Carlo临界计算全堆芯模拟面临几个颇具挑战性的困难,其中包括收敛速度慢、缺乏有效的收敛性判据和计算结果的不确定度被低估等几个方面.本文介绍了Monte Carlo临界计算的过程和特点、存在的困难以及研究现状,综述了最新的加速收敛方法和降方差技巧.The core problem of criticality calculations in nuclear analysis is to calculate the fission source distributions of systems.As the computing performance of hardware and software has advanced,Monte Carlo methods have been applied to the nuclear analysis of whole core problems,because the accuracy of the Monte Carlo calculations has been enhanced by its ability to use continuous energy nuclear data and to handle complex geometry information.When approaching whole core analyses,the Monte Carlo criticality calculations must handle some challenging problems,such as slow convergence,the stopping criteria of source convergence,and real variance estimations.Here,the theory and challenges of Monte Carlo criticality are introduced,and progress in convergence acceleration methods and variance reduction techniques is reviewed.

关 键 词：反应堆分析 临界计算 裂变源分布 MONTE Carlo 中子输运方程 

分 类 号：TL329.2[核科学技术—核技术及应用]