Meta-Auto-Decoder: A Meta-Learning Based Reduced Order Model for Solving Parametric Partial Differential Equations

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Ye, Zhanhong, Huang, Xiang, Liu, Hongsheng, Dong, Bin
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909091588210688
author Ye, Zhanhong
Huang, Xiang
Liu, Hongsheng
Dong, Bin
author_facet Ye, Zhanhong
Huang, Xiang
Liu, Hongsheng
Dong, Bin
contents Many important problems in science and engineering require solving the so-called parametric partial differential equations (PDEs), i.e., PDEs with different physical parameters, boundary conditions, shapes of computational domains, etc. Typical reduced order modeling techniques accelarate solution of the parametric PDEs by projecting them onto a linear trial manifold constructed in the offline stage. These methods often need a predefined mesh as well as a series of precomputed solution snapshots, andmay struggle to balance between efficiency and accuracy due to the limitation of the linear ansatz. Utilizing the nonlinear representation of neural networks, we propose Meta-Auto-Decoder (MAD) to construct a nonlinear trial manifold, whose best possible performance is measured theoretically by the decoder width. Based on the meta-learning concept, the trial manifold can be learned in a mesh-free and unsupervised way during the pre-training stage. Fast adaptation to new (possibly heterogeneous) PDE parameters is enabled by searching on this trial manifold, and optionally fine-tuning the trial manifold at the same time. Extensive numerical experiments show that the MAD method exhibits faster convergence speed without losing accuracy than other deep learning-based methods.
format Preprint
id arxiv_https___arxiv_org_abs_2302_08263
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Meta-Auto-Decoder: A Meta-Learning Based Reduced Order Model for Solving Parametric Partial Differential Equations
Ye, Zhanhong
Huang, Xiang
Liu, Hongsheng
Dong, Bin
Numerical Analysis
68T07
Many important problems in science and engineering require solving the so-called parametric partial differential equations (PDEs), i.e., PDEs with different physical parameters, boundary conditions, shapes of computational domains, etc. Typical reduced order modeling techniques accelarate solution of the parametric PDEs by projecting them onto a linear trial manifold constructed in the offline stage. These methods often need a predefined mesh as well as a series of precomputed solution snapshots, andmay struggle to balance between efficiency and accuracy due to the limitation of the linear ansatz. Utilizing the nonlinear representation of neural networks, we propose Meta-Auto-Decoder (MAD) to construct a nonlinear trial manifold, whose best possible performance is measured theoretically by the decoder width. Based on the meta-learning concept, the trial manifold can be learned in a mesh-free and unsupervised way during the pre-training stage. Fast adaptation to new (possibly heterogeneous) PDE parameters is enabled by searching on this trial manifold, and optionally fine-tuning the trial manifold at the same time. Extensive numerical experiments show that the MAD method exhibits faster convergence speed without losing accuracy than other deep learning-based methods.
title Meta-Auto-Decoder: A Meta-Learning Based Reduced Order Model for Solving Parametric Partial Differential Equations
topic Numerical Analysis
68T07
url https://arxiv.org/abs/2302.08263