Sequencing Initial Conditions in Physics-Informed Neural Networks

Author:

Hooshyar SamanORCID,Elahi Arash

Abstract

The scientific machine learning (SciML) field has introduced a new class of models called physics-informed neural networks (PINNs). These models incorporate domain-specific knowledge as soft constraints on a loss function and use machine learning techniques to train the model. Although PINN models have shown promising results for simple problems, they are prone to failure when moderate level of complexities are added to the problems. We demonstrate that the existing baseline models, in particular PINN and evolutionary sampling (Evo), are unable to capture the solution to differential equations with convection, reaction, and diffusion operators when the imposed initial condition is non-trivial. We then propose a promising solution to address these types of failure modes. This approach involves coupling Curriculum learning with the baseline models, where the network first trains on PDEs with simple initial conditions and is progressively exposed to more complex initial conditions. Our results show that we can reduce the error by 1 – 2 orders of magnitude with our proposed method compared to regular PINN and Evo.

Publisher

Science Research Publishers

Reference51 articles.

1. Acharya N. et al. Instability analysis of poiseuille flow between two parallel walls partially obstructed by porous surfaces. In APS Division of Fluid Dynamics Meeting Abstracts, 2019, pp. NP05-140.

2. Ainsworth M., Dong J., 2021. Galerkin neural networks: a framework for approximating variational equations with error control. SIAM J. Sci. Comput., 43, A2474-A2501. https://doi.org/10.1137/20M1366587

3. Aygün A. Physics informed neural networks for computational fluid dynamics. Middle East Technical University, Master's thesis, 2023.

4. Batra R. et al. Emerging Materials Intelligence Ecosystems Propelled by Machine Learning. Nature Reviews. Materials, 2021, 6(8), 655-678. https://doi.org/10.1038/s41578-020-00255-y

5. Beck C., Hutzenthaler M., Jentzen A., Kuckuck B.J., 2020. An overview on deep learning-based approximation methods for partial differential equations.

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3