Point Cloud Modeling Method for Deep Learning Numerical Calculation of Nuclear Reactor Core

Li Manyuan; Liu Yanli; Liu Dong; Lyu Hengye; Mu Qiao; An Ping; Xing Guanyu; Yang Hongyu; Tu Xiaolan; Pang Zhixin

doi:10.13832/j.jnpe.2024.090048

Volume 46 Issue 5

Oct. 2025

Turn off MathJax

Article Contents

Article Navigation > Nuclear Power Engineering > 2025 > 46(5): 249-257

Li Manyuan, Liu Yanli, Liu Dong, Lyu Hengye, Mu Qiao, An Ping, Xing Guanyu, Yang Hongyu, Tu Xiaolan, Pang Zhixin. Point Cloud Modeling Method for Deep Learning Numerical Calculation of Nuclear Reactor Core[J]. Nuclear Power Engineering, 2025, 46(5): 249-257. doi: 10.13832/j.jnpe.2024.090048

Citation:

Li Manyuan, Liu Yanli, Liu Dong, Lyu Hengye, Mu Qiao, An Ping, Xing Guanyu, Yang Hongyu, Tu Xiaolan, Pang Zhixin. Point Cloud Modeling Method for Deep Learning Numerical Calculation of Nuclear Reactor Core[J]. Nuclear Power Engineering, 2025, 46(5): 249-257. doi: 10.13832/j.jnpe.2024.090048

Citation:

Li Manyuan, Liu Yanli, Liu Dong, Lyu Hengye, Mu Qiao, An Ping, Xing Guanyu, Yang Hongyu, Tu Xiaolan, Pang Zhixin. Point Cloud Modeling Method for Deep Learning Numerical Calculation of Nuclear Reactor Core[J]. Nuclear Power Engineering, 2025, 46(5): 249-257. doi: 10.13832/j.jnpe.2024.090048

PDF( 24943 KB)

Point Cloud Modeling Method for Deep Learning Numerical Calculation of Nuclear Reactor Core

doi: 10.13832/j.jnpe.2024.090048

1.
School of Computer Science, Sichuan University, Chengdu, 610065, China
2.
Nuclear Power Institute of China, Chengdu, 610213, China

Received Date: 2024-09-18
Rev Recd Date: 2024-10-09

Available Online: 2025-10-15

Publish Date: 2025-10-15

Abstract

Abstract

With the development of deep learning technology, the application of deep learning computational methods to solve multi-disciplinary equations in reactor physics, thermal hydraulics, and other fields has become a hot research topic and an important future direction in reactor numerical computation. To address the challenge that traditional mesh structures cannot be directly used for nuclear reactor simulation, this paper investigates a point cloud-based modeling method for nuclear reactor cores at three levels: fuel cell, assembly, and core. The study achieves hierarchical modeling from fuel cells to assemblies and finally to the core, improving the reusability of model data. Based on this modeling approach, a point cloud-based reactor modeling software is developed, featuring a series of functions such as nuclear reactor structure design, point cloud sampling, boundary separation, and attribute visualization. This work represents the first publicly documented reactor modeling method tailored for deep learning numerical computation, providing effective and reusable nuclear reactor data for deep learning simulations. Using real reactor core data, this paper validates the constructed core model through deep learning numerical simulations, confirming the correctness and effectiveness of the models generated by the software.
- Nuclear reactor core,
- Deep learning numerical computation,
- 3D point cloud,
- Nuclear reactor modeling,
- Point cloud visualization

FullText(HTML)

References(18)

References

[1]	XIE Y C, WANG Y H, MA Y, et al. Neural network based deep learning method for multi-dimensional neutron diffusion problems with novel treatment to boundary[J]. Journal of Nuclear Engineering, 2021, 2(4): 533-552. doi: 10.3390/jne2040036
[2]	HERNÁNDEZ A R, GÓMEZ-TORRES A M, DEL VALLE-GALLEGOS E. Nuclear reactor simulation[M]//AWWAD N, ALFAIFY S A. New Trends in Nuclear Science. London: IntechOpen, 2018.
[3]	SJENITZER B L, HOOGENBOOM J E. Dynamic Monte Carlo method for nuclear reactor kinetics calculations[J]. Nuclear Science and Engineering, 2013, 175(1): 94-107. doi: 10.13182/NSE12-44
[4]	SJENITZER B L, HOOGENBOOM J E. A Monte Carlo method for calculation of the dynamic behaviour of nuclear reactors[J]. Progress in Nuclear Science and Technology, 2011, 2: 716-721. doi: 10.15669/pnst.2.716
[5]	CHAULIAC C, ARAGONÉS J M, BESTION D, et al. NURESIM–A European simulation platform for nuclear reactor safety: multi-scale and multi-physics calculations, sensitivity and uncertainty analysis[J]. Nuclear Engineering and Design, 2011, 241(9): 3416-3426. doi: 10.1016/j.nucengdes.2010.09.040
[6]	GASTON D R, PERMANN C J, PETERSON J W, et al. Physics-based multiscale coupling for full core nuclear reactor simulation[J]. Annals of Nuclear Energy, 2015, 84: 45-54. doi: 10.1016/j.anucene.2014.09.060
[7]	Fletcher C D, Schultz R R. RELAP5/MOD3 code manual[R]. Washington, DC (United States): Nuclear Regulatory Commission; Div. of Systems Research; EG and G Idaho, Inc., Idaho Falls, ID (United States), 1992.
[8]	Barre F, Bernard M. The CATHARE code strategy and assessment[J]. Nuclear Engineering and Design, 1990, 124(3): 257-284.
[9]	汤琪芬, 芦韡, 刘婷, 等. 核反应堆堆芯数值计算软件的渐增式集成测试技术研究与应用[J]. 科技视界, 2022(18): 41-44.
[10]	芦韡, 冯晋涛, 潘俊杰. HDF5数据格式在核反应堆计算软件中的应用研究[J]. 中国核电, 2018, 11(4): 475-480.
[11]	王亚曦. 接触分析在反应堆结构设计中的应用[J]. 核动力工程, 2013, 34(S1): 164-167.
[12]	曹国海, 芦韡, 张尧毅, 等. 基于核反应堆数值软件运行特性的并行调度技术研究[J]. 科技视界, 2021(17): 50-53.
[13]	明平洲, 潘俊杰, 芦韡, 等. 全堆芯子通道分析的结构化网格并行算法[J]. 计算机应用, 2017, 37(S2): 35-38.
[14]	刘东, 罗琦, 唐雷, 等. 基于PINN深度机器学习技术求解多维中子学扩散方程[J]. 核动力工程, 2022, 43(2): 1-8.
[15]	刘东, 唐雷, 安萍, 等. 核反应堆有效增殖系数深度学习直接搜索求解方法[J]. 核动力工程, 2023, 44(5): 6-14.
[16]	KEHRER J, HAUSER H. Visualization and visual analysis of multifaceted scientific data: a survey[J]. IEEE Transactions on Visualization and Computer Graphics, 2013, 19(3): 495-513. doi: 10.1109/TVCG.2012.110
[17]	Sievert C. Interactive web-based data visualization with R, plotly, and shiny[M]. New York: Chapman and Hall/CRC, 2020: 1-470.
[18]	LEFF A, RAYFIELD J T. Web-application development using the model/view/controller design pattern[C]//Proceedings of the 5th IEEE International Enterprise Distributed Object Computing Conference. Seattle: IEEE, 2001: 118-127.