Enforcing Dirichlet Boundary Conditions in Operator Learning

DOI10.48550/arXiv.2608.27256MaRDI QIDQ7285071zbMATH OpenarXivFDO

Authors A. M. Stuart, Andrew M. Stuart, Margaret Trautner

Publication date 27 August 2026

Published in arXiv

Tldr The paper presents a novel neural operator architecture that enforces homogeneous Dirichlet boundary conditions independently of training, without imposing impractical restrictions such as smooth boundaries or separable domains, while retaining expressivity of existing kernel-integral architectures.

Full work available at URL https://doi.org/10.48550/arXiv.2608.27256

arXiv classification

cs.LGmath.NA

Cites work

• A Novel Method for Enforcing Exactly Dirichlet, Neumann and Robin Conditions on Curved Domain Boundaries for Physics Informed Machine Learning
• Adaptive approximation and generalization of deep neural network with intrinsic dimensionality
• Das asymptotische Verteilungsgesetz der Eigenwerte linearer partieller Differentialgleichungen (mit einer Anwendung auf die Theorie der Hohlraumstrahlung).
• Data-driven stochastic closure modeling via conditional diffusion model and neural operator
• DeepOKAN: deep operator network based on Kolmogorov Arnold networks for mechanics problems
• Elliptic partial differential equations of second order
• Enforcing Dirichlet boundary conditions in physics-informed neural networks and variational physics-informed neural networks
• Fourier Neural Operator for Parametric Partial Differential Equations
• Global Convergence of Deep Galerkin and PINNs Methods for Solving Partial Differential Equations
• Guiding continuous operator learning through Physics-based boundary constraints
• Learning Memory and Material Dependent Constitutive Laws
• Learning physical models that can respect conservation laws
• Learning with boundary conditions
• (SOFTWARE) margaret
• Model reduction and neural networks for parametric PDEs
• Neural Q-learning for solving PDEs
• Non-intrusive reduced order modeling of nonlinear problems using neural networks
• Nonlocality and nonlinearity implies universality in operator learning
• Physics-Informed Deep Neural Operator Networks
• SPFNO: Spectral operator learning for PDEs with Dirichlet and Neumann boundary conditions
• The Random Feature Model for Input-Output Maps between Banach Spaces
• Ultracontractivity and the heat kernel for Schrödinger operators and Dirichlet Laplacians
• Unified form language: a domain-specific language for weak formulations of partial differential equations
• Wavelet neural operator for solving parametric partial differential equations in computational mechanics problems
• Weyl asymptotic formula for the Laplacian on domains with rough boundaries

Cited in

(1)

• Cost-Accuracy Trade-offs: Neural Operator vs Classical Numerical Solver

This page was built for publication: Enforcing Dirichlet Boundary Conditions in Operator Learning

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7285071)