Abstract:Neural operators, which learn mappings between the function spaces, have been applied to solve boundary value problems in various ways, including learning mappings from the space of the forcing terms to the space of the solutions with the substantial requirements of data pairs. In this work, we present a data-free neural operator integrated with physics, which learns the Green kernel directly. Our method proceeds in three steps: 1. The governing equations for the Green's function are reformulated into an interface problem, where the delta Dirac function is removed; 2. The interface problem is embedded in a lifted space of higher-dimension to handle the jump in the derivative, but still solved on a two-dimensional surface without additional sampling cost; 3. Deep neural networks are employed to address the curse of dimensionality caused by this lifting operation. The approximate Green's function obtained through our approach is then used to construct preconditioners for the linear systems allowed by its mathematical properties. Furthermore, the spectral bias of it revealed through both theoretical analysis and numerical validation contrasts with the smoothing effects of traditional iterative solvers, which motivates us to propose a hybrid iterative method that combines these two solvers. Numerical experiments demonstrate the effectiveness of our approximate Green's function in accelerating iterative methods, proving fast convergence for solving indefinite problems even involving discontinuous coefficients.

Acceleration of convergence in approximate solutions of Urysohn integral equations with Green's kernels

Optimization of Approximate Solution of Integral Equations of Several Variables

Average Optimization of the Approximate Solution of Operator Equations and Its Application

A note on improvement by iteration for the approximate solutions of second kind Fredholm integral equations with Green's kernels

Convergence Behavior of Gauss-Newton's Method and Extensions of the Smale Point Estimate Theory.

Neural Green's Function Accelerated Iterative Methods for Solving Indefinite Boundary Value Problems

Korovkin-type approximation for non positive operators

Fast integral equation methods for linear and semilinear heat equations in moving domains

Fredholm integral equations with non-smooth kernels in weighted spaces: Nyström approximations, stability and convergence

Smale's α-theory for inexact Newton methods under the γ-condition

Superconvergent spectral projection and multi-projection methods for nonlinear Fredholm integral equations

Asymptotic expansions for approximate solutions of boundary integral equations

Approximation of optimization problems with constraints through kernel Sum-Of-Squares

Numerical solutions of Troesch's problem based on a faster iterative scheme with an application

Some approximation results for mild solutions of stochastic fractional order evolution equations driven by Gaussian noise

A novel fixed point approach based on Green's function for solution of fourth order BVPs

A comparative study of discretization techniques for augmented Urysohn type nonlinear functional Volterra integral equations and their convergence analysis

Multi- and Infinite-variate Integration and $L^2$-Approximation on Hilbert Spaces with Gaussian Kernels

A Fast Numerical Solution Method for Two Dimensional Fredholm Integral Equations of the Second Kind

An efficient iterative procedure in hyperbolic space and application to non-linear delay integral equation