Multi-resolution partial differential equations preserved learning framework for spatiotemporal dynamics

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Xin-Yang, Zhu, Min, Lu, Lu, Sun, Hao, Wang, Jian-Xun
Format: Preprint
Published: 2022
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929209039912960
author Liu, Xin-Yang
Zhu, Min
Lu, Lu
Sun, Hao
Wang, Jian-Xun
author_facet Liu, Xin-Yang
Zhu, Min
Lu, Lu
Sun, Hao
Wang, Jian-Xun
contents Traditional data-driven deep learning models often struggle with high training costs, error accumulation, and poor generalizability in complex physical processes. Physics-informed deep learning (PiDL) addresses these challenges by incorporating physical principles into the model. Most PiDL approaches regularize training by embedding governing equations into the loss function, yet this depends heavily on extensive hyperparameter tuning to weigh each loss term. To this end, we propose to leverage physics prior knowledge by ``baking'' the discretized governing equations into the neural network architecture via the connection between the partial differential equations (PDE) operators and network structures, resulting in a PDE-preserved neural network (PPNN). This method, embedding discretized PDEs through convolutional residual networks in a multi-resolution setting, largely improves the generalizability and long-term prediction accuracy, outperforming conventional black-box models. The effectiveness and merit of the proposed methods have been demonstrated across various spatiotemporal dynamical systems governed by spatiotemporal PDEs, including reaction-diffusion, Burgers', and Navier-Stokes equations.
format Preprint
id arxiv_https___arxiv_org_abs_2205_03990
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Multi-resolution partial differential equations preserved learning framework for spatiotemporal dynamics
Liu, Xin-Yang
Zhu, Min
Lu, Lu
Sun, Hao
Wang, Jian-Xun
Machine Learning
Computational Physics
Traditional data-driven deep learning models often struggle with high training costs, error accumulation, and poor generalizability in complex physical processes. Physics-informed deep learning (PiDL) addresses these challenges by incorporating physical principles into the model. Most PiDL approaches regularize training by embedding governing equations into the loss function, yet this depends heavily on extensive hyperparameter tuning to weigh each loss term. To this end, we propose to leverage physics prior knowledge by ``baking'' the discretized governing equations into the neural network architecture via the connection between the partial differential equations (PDE) operators and network structures, resulting in a PDE-preserved neural network (PPNN). This method, embedding discretized PDEs through convolutional residual networks in a multi-resolution setting, largely improves the generalizability and long-term prediction accuracy, outperforming conventional black-box models. The effectiveness and merit of the proposed methods have been demonstrated across various spatiotemporal dynamical systems governed by spatiotemporal PDEs, including reaction-diffusion, Burgers', and Navier-Stokes equations.
title Multi-resolution partial differential equations preserved learning framework for spatiotemporal dynamics
topic Machine Learning
Computational Physics
url https://arxiv.org/abs/2205.03990