arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.05341math.NAcs.NA
收藏

椭圆型偏微分方程的基于残差的有限元代理求解器

A residual-based finite element surrogate solver for elliptic partial differential equations

Ajou University(亚洲大学);Chung-Ang University(中央大学)

Kyoungjin Jung, Jae Yong Lee, Dongwook Shin

AI总结:

提出一种无数据、结合CNN与有限元残差的代理求解器,通过训练损失与H1半范数误差的关系保证最优收敛,并引入分解策略提高效率。

摘要:

我们提出了一种基于残差的有限元代理求解器,用于椭圆型偏微分方程。该方法在无数据设置下将卷积神经网络与经典有限元离散相结合,其中损失函数直接由有限元残差定义。这使得无需配对的输入-输出数据即可逼近解算子。该方法的一个关键特点是可以在网格细化下使用标准有限元理论进行分析。我们建立了训练损失与$H^1$半范数误差之间的关系,并推导出确保最优收敛率的训练准则。为了提高效率,我们引入了一种分解策略,将不同输入分量的贡献分离。这使得模型能够学习更简单的子算子。数值实验表明,所提方法在网格细化下实现稳定收敛,在复杂几何和振荡解上保持鲁棒性,并自然推广到非线性方程。

英文摘要:

We propose a residual-based finite element surrogate solver for elliptic partial differential equations. The method combines convolutional neural networks with classical finite element discretization in a data-free setting, where the loss function is defined directly from the finite element residual. This enables the approximation of the solution operator without requiring paired input-output data. A key feature of the proposed approach is that it can be analyzed using standard finite element theory under mesh refinement. We establish a relationship between the training loss and the error in the $H^1$-seminorm, and derive training criteria that ensure optimal convergence rates. To improve efficiency, we introduce a decomposition strategy that separates the contributions of different input components. This allows the model to learn simpler sub-operators. Numerical experiments demonstrate that the proposed method achieves stable convergence under grid refinement, remains robust on complex geometries and oscillatory solutions, and extends naturally to nonlinear equations.

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。

↑