Research on post-processing methods for particle radiation Monte Carlo calculations of large-scale target objects
-
摘要: 蒙特卡罗（Monte Carlo，MC）方法是辐照损伤、辐照屏蔽研究中应用最广泛的方法之一。在对机场、铁路、舰船等大型靶目标开展辐照损伤研究时，通常关注靶目标的3D建模及辐照计算，而对计算后的数据分析多采用人工方式，工作难度大、效率低，成为制约相关研究的技术瓶颈。开展靶目标粒子辐照MC计算可视化后处理方法研究，建立了基于KD树（k-dimensional tree，简称KDtree）+反距离加权（inverse distance weight，IDW）和基于遗传算法优化反向传播（genetic algorithm based backpropagation，GABP）神经网络的后处理模型，实现了数据与模型结合的可视化分析。与传统数据分析方法相比，提出的方法能够大幅减低研究人员工作难度，提升数据处理速度，实现辐照效应直观展示，提升辐照效应研究后处理工作效率。Abstract: The Monte Carlo (MC) method is one of the most widely applied methods in the simulation study of radiation damage and radiation shielding. When conducting radiation damage studies on large targets such as airports, railways, and ships, the focus is generally on 3D modeling and radiation calculations of these targets; however, the post-calculation data analysis often relies on manual methods, making this aspect of the research technically challenging and inefficient, thus becoming a bottleneck in related research efforts. In this paper, a visualization post-processing method for MC calculations of target particle irradiation is studied, and a post-processing model based on k-dimensional tree (KDtree) + inverse distance weighting (IDW) and genetic algorithm based backpropagation (GABP) neural network is established to realize the visualization analysis of data combined with the model. Compared with traditional data analysis methods, the method proposed in this paper can greatly reduce the difficulty of researchers’ work, improve the speed of data processing, realize the visual display of radiation effects, and enhance the efficiency of post-processing in radiation effects research.
-
Key words:
- particle irradiation /
- genetic algorithm /
- BP neural network /
- KDtree /
- inverse distance weighting /
- post-processing
-
图 1 MC模型代码可视化截面图
Figure 1. Visualization of cross-sectional diagrams for MC model code
图 2 中子球源及网格计数区域示意图
Figure 2. Schematic diagram of neutron source and grid counting area
图 3 基于KD树+IDW的MC计算后处理模型
Figure 3. Post-processing model for MC calculation based on KDtree+IDW
图 5 基于GABP神经网络的MC计算后处理模型
Figure 5. Post-processing model for MC calculation based on GABP neural network
图 6 后处理模型预测数据与原始数据对比
Figure 6. Comparison of post-processed model prediction data with original data
图 8 基于MC网格计数中子通量空间分布
Figure 8. Spatial distribution of neutron flux counting based on MC grid
表 1 模型性能评估
Table 1. Model performance evaluation
model R2 MSE RMSE MAE location 1 KDtree+IDW model 0.997 1 0.000 39 0.019 66 0.109 22 BP model 0.995 8 0.000 92 0.030 33 0.144 33 GABP model 0.995 5 0.000 92 0.030 43 0.125 97 location 2 KDtree+IDW model 0.999 0 0.000 14 0.011 73 0.088 28 BP model 0.998 3 0.000 51 0.022 59 0.131 06 GABP model 0.998 5 0.000 43 0.020 87 0.127 51 
下载: 导出CSV
-
[1] 梁金刚, 王侃, 蔡云, 等. 中子输运蒙特卡罗模拟的区域分解方法研究[J]. 原子能科学技术, 2014, 48(12):2315-2320 doi: 10.7538/yzk.2014.48.12.2315Liang Jingang, Wang Kan, Cai Yun, et al. Study on domain decomposition method for Monte Carlo simulation of neutron transport[J]. Atomic Energy Science and Technology, 2014, 48(12): 2315-2320 doi: 10.7538/yzk.2014.48.12.2315 [2] Heranudin, Smith M L, Van Wyngaardt W M, et al. Characterisation of a well-type NaI(Tl) detector by means of a Monte Carlo simulation for radionuclide metrology application[J]. Applied Radiation and Isotopes, 2021, 176: 109889. doi: 10.1016/j.apradiso.2021.109889 [3] 马锐垚, 王鑫, 李树, 等. 基于神经网络的粒子输运问题高效计算方法[J]. 物理学报, 2024, 73:072802 doi: 10.7498/aps.73.20231661Ma Ruiyao, Wang Xin, Li Shu, et al. An efficient calculation method for particle transport problems based on neural network[J]. Acta Physica Sinica, 2024, 73: 072802 doi: 10.7498/aps.73.20231661 [4] 李静文, 孙国民, 周文, 等. 蒙特卡罗仿真计算前处理平台的开发与初步验证[J]. 核安全, 2023, 22(1):81-88 doi: 10.3969/j.issn.1672-5360.2023.1.haq202301014Li Jingwen, Sun Guomin, Zhou Wen, et al. Development and preliminary verification of Monte Carlo pre-processing platform[J]. Nuclear Safety, 2023, 22(1): 81-88 doi: 10.3969/j.issn.1672-5360.2023.1.haq202301014 [5] Harb M, Leichtle D, Fischer U. A novel algorithm for CAD to CSG conversion in McCAD[J]. Journal of Nuclear Engineering, 2023, 4(2): 436-447. doi: 10.3390/jne4020031 [6] 陈念年, 蔡勇, 张建生, 等. MCNP输出数据分析与可视化软件的设计与实现[J]. 计算机工程与应用, 2010, 46(8):208-210,231 doi: 10.3778/j.issn.1002-8331.2010.08.060Chen Niannian, Cai Yong, Zhang Jiansheng, et al. Design and realization of MCNP output data analysis and visualization software[J]. Computer Engineering and Applications, 2010, 46(8): 208-210,231 doi: 10.3778/j.issn.1002-8331.2010.08.060 [7] 范佳锦, 王义, 程建平, 等. MCNP4B程序可视化运行平台开发[J]. 核电子学与探测技术, 2002, 22(1):52-55 doi: 10.3969/j.issn.0258-0934.2002.01.015Fan Jiajin, Wang Yi, Cheng Jianping, et al. Development of visual platform of MCNP4B[J]. Nuclear Electronics & Detection Technology, 2002, 22(1): 52-55 doi: 10.3969/j.issn.0258-0934.2002.01.015 [8] 黄燕, 唐登攀, 陈法新. MCNP输入、输出接口的设计[J]. 电子测量与仪器学报, 2007, 21(1):102-106Huang Yan, Tang Dengpan, Chen Faxin. MCNP input / output interface[J]. Journal of Electronic Measurement and Instrument, 2007, 21(1): 102-106 [9] Huang Yongheng, Kim J, Moreira R G, et al. A web-based information system for MCNP simulation of irradiation of complex shaped foods[J]. Applied Engineering in Agriculture, 2008, 24(2): 233-242. [10] 蓝岭. “尼米兹”级航母抗毁伤能力深化研究[J]. 鱼雷技术, 2002, 10(3):8-13Lan Ling. The deepened research in the anti-damaged capability of the Nimitz class aircraft carrier[J]. Torpedo Technology, 2002, 10(3): 8-13 [11] 徐峰, 周心桃, 李德聪, 等. 国外典型航母目标特性探析[J]. 舰船科学技术, 2018, 40(17):151-157 doi: 10.3404/j.issn.1672-7649.2018.09.030Xu Feng, Zhou Xintao, Li Decong, et al. Analysis of the target characteristics of foreign typical aircraft carrier[J]. Ship Science and Technology, 2018, 40(17): 151-157 doi: 10.3404/j.issn.1672-7649.2018.09.030 [12] Lu Lei, Qiu Yuefeng, Fischer U. Improved solid decomposition algorithms for the CAD-to-MC conversion tool McCad[J]. Fusion Engineering and Design, 2017, 124: 1269-1272. doi: 10.1016/j.fusengdes.2017.02.040 [13] Lu L, Fischer U, Pereslavtsev P. Improved algorithms and advanced features of the CAD to MC conversion tool McCad[J]. Fusion Engineering and Design, 2014, 89(9/10): 1885-1888. [14] Moro F, Fischer U, Lu Lei, et al. The McCad code for the automatic generation of MCNP 3-D models: applications in fusion neutronics[J]. IEEE Transactions on Plasma Science, 2014, 42(4): 1036-1041. doi: 10.1109/TPS.2014.2308957 [15] 康美泽, 马英华. 美国航母结构用钢情况分析[J]. 船舶标准化与质量, 2016(1):46-48,54Kang Meize, Ma Yinghua. Analysis of steel used in the structure of American aircraft carriers[J]. Shipbuilding Standardization & Quality, 2016(1): 46-48,54 [16] Abella V, Miró R, Juste B, et al. 3D dose distribution calculation in a voxelized human phantom by means of Monte Carlo method[J]. Applied Radiation and Isotopes, 2010, 68(4/5): 709-713. [17] Zhang Shichao, Li Xuelong, Zong Ming, et al. Learning k for kNN classification[J]. ACM Transactions on Intelligent Systems and Technology (TIST), 2017, 8: 43. [18] 夏军勇, 高睿杰, 钟飞. 基于KDTree改进的Super-4PCS+ICP算法在点云配准中的应用研究[J]. 激光与红外, 2023, 53(9):1333-1338 doi: 10.3969/j.issn.1001-5078.2023.09.005Xia Junyong, Gao Ruijie, Zhong Fei. Study on the application of Super-4 PCS+ICP algorithm based on KDTree improvement in point cloud alignment[J]. Laser & Infrared, 2023, 53(9): 1333-1338 doi: 10.3969/j.issn.1001-5078.2023.09.005 [19] Chen C S, Karageorghis A. Local RBF algorithms for elliptic boundary value problems in annular domains[J]. Communications in Computational Physics, 2019, 25(1): 41-67. doi: 10.4208/cicp.OA-2018-0031 [20] 廖丽琼, 白俊松, 罗德安. 基于八叉树及KD树的混合型点云数据存储结构[J]. 计算机系统应用, 2012, 21(3):87-90 doi: 10.3969/j.issn.1003-3254.2012.03.020Liao Liqiong, Bai Junsong, Luo Dean. Integrated point cloud storage structure based on octree and KDTree[J]. Computer Systems & Applications, 2012, 21(3): 87-90 doi: 10.3969/j.issn.1003-3254.2012.03.020 [21] 范晶晶, 王力, 褚文博, 等. 基于KDTree树和欧式聚类的越野环境下行人识别的研究[J]. 汽车工程, 2019, 41(12):1410-1415Fan Jingjing, Wang Li, Chu Wenbo, et al. Research on pedestrian recognition in cross-country environment based on KDTree and Euclidean clustering[J]. Automotive Engineering, 2019, 41(12): 1410-1415 [22] 蒋厚伟, 陈宗玉, 刘德亮, 等. 基于声发射和GA-BP神经网络的铣削SiCp/Al表面粗糙度预测[J]. 软件工程, 2024, 27(4):6-12Jiang Houwei, Chen Zongyu, Liu Deliang, et al. Prediction of surface roughness of SiCp/Al milling based on acoustic emission and GA-BP neural network[J]. Software Engineering, 2024, 27(4): 6-12 -
点击查看大图
图(9) / 表(1)
计量
- 文章访问数: 88
- HTML全文浏览量: 32
- PDF下载量: 12
- 被引次数: 0
