Research on Rapid Calculation and Reconstruction Technology of Radiation Field Based on Bayesian Inference Method
-
摘要: 为了满足核设施退役、核电厂维修等相关现场工作辐射场快速计算的需求，本文开发了基于点核积分方法，且适用于复杂几何和源描述的辐射场快速计算软件CSGPK。为了提升CSGPK软件点核积分模块的计算精度，使用贝叶斯推断方法对辐射场进行了重构，并使用平板模型和迷宫结构对CSGPK软件进行了验证。验证结果表明，CSGPK软件点核积分模块的计算精度与国际同类计算软件相当；通过使用贝叶斯推断方法进行辐射场重构，能够大幅提升点核积分方法的计算精度，并提升CSGPK软件的适用范围。Abstract: In order to meet the requirements of rapid calculation of radiation field in nuclear facility decommissioning, nuclear power plant maintenance and other related field work, a rapid calculation code CSGPK based on point kernel integral method is developed in this study, which is suitable for complex geometry and source description. In order to improve the calculation accuracy of point kernel integral method, the radiation field is reconstructed by Bayesian inference method, and the developed code is verified by flat plate model and maze model. The verification results show that the calculation accuracy of CSGPK of point kernel method is equivalent to that of similar international calculation software. By using Bayesian inference method to reconstruct the radiation field, the calculation accuracy of the point kernel integration method can be greatly improved, and the application scope of CSGPK can be improved.
-
Key words:
- Point kernel method /
- Bayesian inference method /
- Radiation field reconstruction /
- CSGPK
-
图 1 CSGPK软件点核积分模块计算流程图
Figure 1. Calculation Flow Chart of CSGPK Point Kernel Integral Code
图 3 CSGPK软件与QAD程序的剂量率计算结果的相对误差
Figure 3. Relative Error of CSGPK Dose Calculation Result and QAD Dose Calculation Result
图 4 CSGPK软件点核积分模块剂量率计算结果与MCNP程序直接计算结果的相对误差
Figure 4. Relative Error of Dose Calculation Results between CSGPK Point Kernel Integration Module and MCNP Direct Calculation
图 5 重构后的辐射场与MCNP程序直接计算剂量率的相对误差
Figure 5. Relative Error of Dose Calculation Results between Reconstructed Radiation Field and MCNP Direct Calculation
图 6 迷宫结构CSGPK软件计算得到的辐射场分布
Figure 6. Radiation Field Distribution Calculated by CSGPK for a Maze Structure
图 7 迷宫结构CSGPK软件剂量率计算结果与MCNP程序计算结果的相对误差
Figure 7. Relative Error of Dose Calculation Results between CSGPK and MCNP for a Maze Structure
-
[1] 宋英明,梁烨,叶凯萱,等. 核设施退役过程中的辐射场重构与拆除路径优化[J]. 核技术,2017, 40(5): 050502. [2] ZAZULA J M. BALTORO: a general purpose code for coupling of Monte Carlo and discrete ordinates radiation transport calculations[J] Nuclear Science and Engineering, 2017, 87(2): 196-197. [3] SAFIGHOLI H, SONG W Y. Calculation of water equivalent ratios for various materials at proton energies ranging 10-500 MeV using MCNP, FLUKA, and GEANT4 Monte Carlo codes[J]. Physics in Medicine & Biology, 2018, 63(15): 155010. [4] 胡二邦,高占荣. 广东大亚湾核电站周围建筑物辐射屏蔽因子的计算[J]. 辐射防护,2003, 23(2): 74-83. doi: 10.3321/j.issn:1000-8187.2003.02.002 [5] 李华,赵原,曹勤剑,等. 基于源活度反演计算的γ辐射场数据重构方法[J]. 清华大学学报(自然科学版),2020, 60(10): 880-886. [6] 谢明亮,杨森权,彭波,等. 基于体素化点核算法的三维辐射场计算应用研究[J]. 核动力工程,2020, 41(S1): 82-86. [7] 包彦,贺清明,曹良志,等. 基于NECP-MCX的点核积分方法研究与程序研发[C]. 北京: 中国核科学技术进展报告(第七卷)——中国核学会2021年学术年会论文集,第6册,2021,10. [8] PENG X J, QING D. Reactor core power mapping based on Bayesian inference[J]. Annals of Nuclear Energy, 2018, 119: 322-330. doi: 10.1016/j.anucene.2018.05.014 [9] WANG Z, CAI J J. Reconstruction of the neutron radiation field on nuclear facilities near the shield using Bayesian inference[J]. Progress in Nuclear Energy, 2020, 118: 103070. doi: 10.1016/j.pnucene.2019.103070 -
下载:
计量
- 文章访问数: 32
- HTML全文浏览量: 15
- PDF下载量: 12
- 被引次数: 0
