arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2105.05690math.NAcs.LGcs.NAphysics.comp-ph
收藏

辐射传输方程的机器学习矩闭合模型I:直接学习基于梯度的闭合

Machine learning moment closure models for the radiative transfer equation I: directly learning a gradient based closure

Michigan State University(密歇根州立大学);National Superconducting Cyclotron Laboratory(国家超导回旋加速器实验室)

Juntao Huang, Yingda Cheng, Andrew J. Christlieb, Luke F. Roberts

摘要:

本文采用数据驱动的方法,将机器学习应用于平板几何中辐射传输方程的矩封闭问题。我们提出用神经网络直接学习高阶矩的梯度,而不是学习未闭合的高阶矩。这种新方法与我们针对自由流限制推导出的精确闭包一致,并且还提供了自然的输出归一化。各种基准测试,包括变量散射问题,周期性和反射边界的高斯信源问题,以及两种材料问题,都显示了我们的机器学习闭合模型的良好准确性和泛化能力。

英文摘要:

In this paper, we take a data-driven approach and apply machine learning to the moment closure problem for radiative transfer equation in slab geometry. Instead of learning the unclosed high order moment, we propose to directly learn the gradient of the high order moment using neural networks. This new approach is consistent with the exact closure we derive for the free streaming limit and also provides a natural output normalization. A variety of benchmark tests, including the variable scattering problem, the Gaussian source problem with both periodic and reflecting boundaries, and the two-material problem, show both good accuracy and generalizability of our machine learning closure model.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑