Abstract:Physics-informed neural networks (PINNs) are appealing data-driven tools for solving and inferring solutions to nonlinear partial differential equations (PDEs). Unlike traditional neural networks (NNs), which train only on solution data, a PINN incorporates a PDE's residual into its loss function and trains to minimize the said residual at a set of collocation points in the solution domain. This paper explores the use of the Schwarz alternating method as a means to couple PINNs with each other and with conventional numerical models (i.e., full order models, or FOMs, obtained via the finite element, finite difference or finite volume methods) following a decomposition of the physical domain. It is well-known that training a PINN can be difficult when the PDE solution has steep gradients. We investigate herein the use of domain decomposition and the Schwarz alternating method as a means to accelerate the PINN training phase. Within this context, we explore different approaches for imposing Dirichlet boundary conditions within each subdomain PINN: weakly through the loss and/or strongly through a solution transformation. As a numerical example, we consider the one-dimensional steady state advection-diffusion equation in the advection-dominated (high Peclet) regime. Our results suggest that the convergence of the Schwarz method is strongly linked to the choice of boundary condition implementation within the PINNs being coupled. Surprisingly, strong enforcement of the Schwarz boundary conditions does not always lead to a faster convergence of the method. While it is not clear from our preliminary study that the PINN-PINN coupling via the Schwarz alternating method accelerates PINN convergence in the advection-dominated regime, it reveals that PINN training can be improved substantially for Peclet numbers as high as 1e6 by performing a PINN-FOM coupling.

The role of interface boundary conditions and sampling strategies for Schwarz-based coupling of projection-based reduced order models

Domain Decomposition-based coupling of Operator Inference reduced order models via the Schwarz alternating method

An explicit substructuring method for overlapping domain decomposition based on stochastic calculus

Schwarz Alternating Method Based on Natural Boundary Reduction for Time-Dependent Problems on Unbounded Domains

Generalized bridging domain method for coupling finite elements with discrete elements

A reduced order model for domain decompositions with non-conforming interfaces

Optimisation-Based Coupling of Finite Element Model and Reduced Order Model for Computational Fluid Dynamics

Domain decomposition-based coupling of physics-informed neural networks via the Schwarz alternating method

Multiscale Mixed Methods with Improved Accuracy: The Role of Oversampling and Smoothing

Scalable Semismooth Newton Methods with Multilevel Domain Decomposition for Subsurface Flow and Reactive Transport in Porous Media

Recent Advances in Multiscale Simulations: Wideband Fast Integral Solver and Domain Decomposition Method

Robust domain decomposition methods for high-contrast multiscale problems on irregular domains with virtual element discretizations

A Generalized Schwarz-type Non-overlapping Domain Decomposition Method using Physics-constrained Neural Networks

Optimal Surrogate Boundary Selection and Scalability Studies for the Shifted Boundary Method on Octree Meshes

Optimized Schwarz Waveform Relaxation for Advection Reaction Diffusion Equations in Two Dimensions

A Domain Decomposition Approach for Local Mesh Refinement in Space and Time

A Schwarz alternating algorithm for elliptic boundary value problems in an infinite domain with a concave angle

Local surrogate models with reduced dimensionality via overlapping domain decomposition and proper generalized decomposition