A Least-Squares-Based Regularity-Conforming Neural Networks (LS-ReCoNNs) for Solving Parametric Transmission Problems
University of the Basque Country (UPV/EHU)(巴斯克大学);CUNEF Universidad(CUNEF大学);Basque Center for Applied Mathematics (BCAM)(巴斯克应用数学中心);Ikerbasque (Basque Foundation For Sciences)(伊克尔巴斯基奎(巴斯克科学基金会))
Shima Baharlouei, Jamie Taylor, David Pardo
备注:
33 pages, 18 figures. Funded by: EU Horizon Europe MSCA-DN-101119556 (IN-DEEP); MICIU/AEI PID2023-146678OB-I00 & CEX2021-001142-S; Basque Govt BERC 2022-2025, ELKARTEK (KK-2023/00012, KK-2024/00086), MATHMODE (IT1866-26); and NextGenerationEU/PRTR
This article focuses on solving parametric transmission problems in one and two spatial dimensions. These problems belong to a class of partial differential equations that arise in the modeling of physical systems with heterogeneous materials. They often exhibit discontinuities across interfaces and singularities at points where interfaces intersect. To address these problems, we propose a new deep learning approach named {\it{Least-Squares-Based Regularity-Conforming Neural Network (LS-ReCoNN)}}. This approach proposes a loss function that is shown to be a consistent upper bound for the energy-norm error. The method represents the solution as the sum of a principal component and a singular component. The principal component is decomposed into smooth and gradient-jump parts, which capture both the regular solution behavior and reduced regularity across interfaces in one- and two-dimensional problems. The singular component is introduced to model junction singularities and it is approximated using basis functions computed from a one-dimensional finite element eigenvalue problem. For the principal component, a separated representation is employed, consisting of parameter-dependent coefficients and space-dependent functions. A deep neural network approximates the space-dependent functions, while the parameter-dependent coefficients are determined by a least-squares solver, where the optimal coefficients for each parameter instance are obtained online by solving a low-dimensional least-squares problem. Numerical experiments in one and two dimensions demonstrate that LS-ReCoNN effectively captures singularities while maintaining solution accuracy across a wide range of parameter values.
Finite element approximation for a delayed generalized Burgers-Huxley equation with weakly singular kernels: Part I Well-posedness, Regularity and Conforming approximation
Conforming and non-conforming virtual element methods for the biharmonic Steklov eigenvalue problem with minimum regularity
Indian Institute of Technology-Kharagpur(印度理工学院卡拉格普尔分校);King Abdullah University of Science and Technology (KAUST)(阿卜杜拉国王科技大学);Università di Pavia(帕维亚大学);IMATI, Consiglio Nazionale delle Ricerche(国家研究委员会IMATI研究所);Universidad Técnica Federico Santa María(费德里科·圣塔玛丽亚技术大学)
Discretisation of heterogeneous and anisotropic diffusion problems on general non-conforming meshes. SUSHI: a scheme using stabilisation and hybrid interfaces
Université Paris-Est(巴黎东大学);Université de Provence(普罗旺斯大学)
Non-Conforming Structure Preserving Finite Element Method for Doubly Diffusive Flows on Bounded Lipschitz Domains
Center for Computation and Technology, Louisiana State University(路易斯安那州立大学计算与技术中心);Department of Mathematics, Indian Institute of Technology, Roorkee(印度理工学院鲁尔基分校数学系)
The gradient discretisation method for optimal control problems, with super-convergence for non-conforming finite elements and mixed-hybrid mimetic finite differences
Monash University(莫纳什大学);Indian Institute of Technology Bombay(印度理工学院孟买分校)
Duality analysis of interior penalty discontinuous Galerkin methods under minimal regularity and application to the a priori and a posteriori error analysis of Helmholtz problems