| Issue |
EPJ Web Conf.
Volume 247, 2021
PHYSOR2020 – International Conference on Physics of Reactors: Transition to a Scalable Nuclear Future
|
|
|---|---|---|
| Article Number | 03013 | |
| Number of page(s) | 8 | |
| Section | Deterministic Transport | |
| DOI | https://doi.org/10.1051/epjconf/202124703013 | |
| Published online | 22 February 2021 | |
https://doi.org/10.1051/epjconf/202124703013
A DEEP LEARNING BASED SURROGATE MODEL FOR ESTIMATING THE FLUX AND POWER DISTRIBUTION SOLVED BY DIFFUSION EQUATION
1 Harbin Engineering University 145 Nantong St, Harbin, Heilongjiang, 150001, China
2 Sino-French Institute of Nuclear Engineering and Technology, Sun Yat-Sen University No.2 Daxue Road, Zhuhai, 519082, China
3 hanghai Jiaotong University 800 Dongchuan Rd. Minhang District, Shanghai, 200240, China
qianzhang@hrbeu.edu.cn
13935397912@hrbeu.edu.cn
liangliang_ls@ hrbeu.edu.cn
lizh333@mail.sysu.edu.cn
zhangtengfei@sjtu.edu.cn
Published online: 22 February 2021
A deep learning based surrogate model is proposed for replacing the conventional diffusion equation solver and predicting the flux and power distribution of the reactor core. Using the training data generated by the conventional diffusion equation solver, a special designed convolutional neural network inspired by the FCN (Fully Convolutional Network) is trained under the deep learning platform TensorFlow. Numerical results show that the deep learning based surrogate model is effective for estimating the flux and power distribution calculated by the diffusion method, which means it can be used for replacing the conventional diffusion equation solver with high efficiency boost.
Key words: deep learning / surrogate model / convolutional neural network / diffusion
© The Authors, published by EDP Sciences, 2021
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.