Implicit Neural Solver for Time-dependent Linear PDEs with Convergence Guarantee

10/08/2019
by   Suprosanna, et al.
1

Fast and accurate solution of time-dependent partial differential equations (PDEs) is of key interest in many research fields including physics, engineering, and biology. Generally, implicit schemes are preferred over the explicit ones for better stability and correctness. The existing implicit schemes are usually iterative and employ a general-purpose solver which may be sub-optimal for a specific class of PDEs. In this paper, we propose a neural solver to learn an optimal iterative scheme for a class of PDEs, in a data-driven fashion. We achieve this goal by modifying an iteration of an existing semi-implicit solver using a deep neural network. Further, we provide theoretical proof that our approach preserves the correctness and convergence guarantees provided by the existing iterative-solvers. We also demonstrate that our model generalizes to a different parameter setting than the one seen during training and achieves faster convergence compared to the semi-implicit schemes.

READ FULL TEXT
research
09/03/2021

Semi-Implicit Neural Solver for Time-dependent Partial Differential Equations

Fast and accurate solutions of time-dependent partial differential equat...
research
06/04/2019

Learning Neural PDE Solvers with Convergence Guarantees

Partial differential equations (PDEs) are widely used across the physica...
research
01/12/2020

High order semi-implicit multistep methods for time dependent partial differential equations

We consider the construction of semi-implicit linear multistep methods w...
research
09/22/2022

A new reduced order model of linear parabolic PDEs

How to build an accurate reduced order model (ROM) for multidimensional ...
research
11/12/2020

A data driven heuristic for rapid convergence of general Scheduled Relaxation Jacobi (SRJ) schemes

The Scheduled Relaxation Jacobi (SRJ) method is a viable candidate as a ...
research
05/26/2023

Stability of implicit neural networks for long-term forecasting in dynamical systems

Forecasting physical signals in long time range is among the most challe...
research
07/10/2023

Galerkin-Bernstein Approximations of the System of Time Dependent Nonlinear Parabolic PDEs

The purpose of the research is to find the numerical solutions to the sy...

Please sign up or login with your details

Forgot password? Click here to reset