第 47 周 10 篇 覆盖区间: 2026-11-06 → 2026-11-13

第 47 周 · 物理信息约束、PINN(物理信息神经网络)等方向的最新进展

本期 10 篇覆盖 物理信息约束、PINN(物理信息神经网络)、PINN(物理信息神经网络) 等方向,共同指向同一个问题:如何让物理约束的学习方法在真实求解尺度上稳定成立。下面按主题逐篇给出导读与推荐理由。

本期要点

  • 1. Fuse-PINN: Combining Finite Element Methods and Physics-Informed Neural Networks to enable massively …(物理信息约束、PINN(物理信息神经网络)方向)
  • 2. CoupledFlow: One-Step Neural Operators for Coupled Multi-Physics PDEs(神经算子、算子学习方向)
  • 3. Convergence analysis of physics-informed neural networks for subdiffusion(物理信息约束、PINN(物理信息神经网络)方向)
  • 4. Parameterized Physics-Informed Neural Networks for Efficient Simulation of Embedded Microfluidic Chip Cooling(物理信息约束、PINN(物理信息神经网络)方向)
  • 本期 10 篇,方向集中在 物理信息约束、PINN(物理信息神经网络);推荐理由基于论文摘要自动生成,详见每条导读

本期收录

1

Fuse-PINN: Combining Finite Element Methods and Physics-Informed Neural Networks to enable massively …

Fuse-PINN: Combining Finite Element Methods and Physics-Informed Neural Networks to enable massively parallel simulations of non-linear PDEs
作者 Stanislas Pamela, Nicolas Pailliez, Emmanuel Franck et al. 9 pages plus Appendix, 11 figures, 1 table. Rejected at ICLR
为什么值得看

探讨了Fuse-PINN: Combining Finite Element Methods Physics-Informed的相关问题,值得关注。

arXiv 原文 · 2611.18409 ↗ physics-informedpinnpinnspde
2

CoupledFlow: One-Step Neural Operators for Coupled Multi-Physics PDEs

CoupledFlow: One-Step Neural Operators for Coupled Multi-Physics PDEs
作者 Trong Khiem Tran, Minh Long Bui, Phi Le Nguyen et al. Accepted NeurIPS 2026
为什么值得看

探讨了CoupledFlow: One-Step Operators Coupled Multi-Physics PDEs的相关问题,值得关注。

arXiv 原文 · 2611.10945 ↗ neural operatoroperator learningpdemulti-physics
3

Convergence analysis of physics-informed neural networks for subdiffusion

Convergence analysis of physics-informed neural networks for subdiffusion
作者 Bangti Jin, Qiqi Rao, Zhi Zhou
为什么值得看

对Convergence analysis physics-informed networks subdiffusion进行了系统性理论分析,给出了收敛性及误差界。

arXiv 原文 · 2611.17759 ↗ physics-informedpinnpinnspartial differential equation
4

Parameterized Physics-Informed Neural Networks for Efficient Simulation of Embedded Microfluidic Chip Cooling

Parameterized Physics-Informed Neural Networks for Efficient Simulation of Embedded Microfluidic Chip Cooling
作者 Weiheng Li, Zhiyao Jiang, Bai Song
为什么值得看

探索了Parameterized Physics-Informed Networks Simulation Embedded Microfluidic在实际场景中的应用效果。

arXiv 原文 · 2611.10666 ↗ physics-informedpinnpinnssurrogate model
5

Rethinking How to Leverage Cheap Physics for Data-Efficient Neural Operators

Rethinking How to Leverage Cheap Physics for Data-Efficient Neural Operators
作者 Shuo Zhang, Senwei Liang
为什么值得看

探讨了Rethinking Leverage Cheap Physics Data-Efficient Operators的相关问题,值得关注。

arXiv 原文 · 2611.10217 ↗ physics-informedneural operatorpde
6

A systematic comparison of physics-informed neural networks and mechanistic models for epidemic …

A systematic comparison of physics-informed neural networks and mechanistic models for epidemic modeling and forecasting
作者 Martina Rama, Yuhan Li, Bruno Lepri et al.
为什么值得看

探讨了systematic comparison physics-informed networks mechanistic models的相关问题,值得关注。

arXiv 原文 · 2611.18389 ↗ physics-informedpinnpinns
7

Well-Balanced WE-PINNs for the Shallow Water Equations over Smooth and Discontinuous Bathymetries

Well-Balanced WE-PINNs for the Shallow Water Equations over Smooth and Discontinuous Bathymetries
作者 Ismail Oubarka, Imad Kissami, Mohamed Boubekeur et al.
为什么值得看

探讨了Well-Balanced WE-PINNs Shallow Water Equations over的相关问题,值得关注。

arXiv 原文 · 2611.17289 ↗ physics-informedpinnpinnspartial differential equation
8

Physics-informed neural networks in the small-data regime: A controlled comparison with classical …

Physics-informed neural networks in the small-data regime: A controlled comparison with classical PDE surrogates
作者 Kirill Katsuba, Oleg Rudenko, Egor Bugaev et al.
为什么值得看

探讨了Physics-informed networks small-data regime: controlled comparison的相关问题,值得关注。

arXiv 原文 · 2611.12424 ↗ physics-informedpinnpinns
9

On the Global Convergence of Stochastic Adam for Physics-Informed Neural Networks

On the Global Convergence of Stochastic Adam for Physics-Informed Neural Networks
作者 Yihang Gao
为什么值得看

对Global Convergence Stochastic Adam Physics-Informed Networks进行了系统性理论分析,给出了收敛性及误差界。

arXiv 原文 · 2611.11905 ↗ physics-informedpinnpinns
10

ALETHEIA: A Multi-Frequency Eddy Current Pulsed Thermography Dataset for Neural Operator Learning …

ALETHEIA: A Multi-Frequency Eddy Current Pulsed Thermography Dataset for Neural Operator Learning in Nondestructive Testing
作者 Changbin Sun, Jie Xiao, Xiaotian Chen et al. 62 pages, 19 figures
为什么值得看

提出一种ALETHEIA: Multi-Frequency Eddy Current Pulsed Thermography的方法,在多个任务上验证了其有效性。

arXiv 原文 · 2611.11252 ↗ neural operatorscientific machine learningpdepartial differential equation

本期正文

本期导读

本期 10 篇覆盖 物理信息约束、PINN(物理信息神经网络)、PINN(物理信息神经网络) 等方向,共同指向同一个问题:如何让物理约束的学习方法在真实求解尺度上稳定成立。下面按主题逐篇给出导读与推荐理由。

本期要点

    1. Fuse-PINN: Combining Finite Element Methods and Physics-Informed Neural Networks to enable massively …(物理信息约束、PINN(物理信息神经网络)方向)
    1. CoupledFlow: One-Step Neural Operators for Coupled Multi-Physics PDEs(神经算子、算子学习方向)
    1. Convergence analysis of physics-informed neural networks for subdiffusion(物理信息约束、PINN(物理信息神经网络)方向)
    1. Parameterized Physics-Informed Neural Networks for Efficient Simulation of Embedded Microfluidic Chip Cooling(物理信息约束、PINN(物理信息神经网络)方向)
  • 本期 10 篇,方向集中在 物理信息约束、PINN(物理信息神经网络);推荐理由基于论文摘要自动生成,详见每条导读

论文推荐

1. Fuse-PINN: Combining Finite Element Methods and Physics-Informed Neural Networks to enable massively …

探讨了Fuse-PINN: Combining Finite Element Methods Physics-Informed的相关问题,值得关注。

来源:9 pages plus Appendix, 11 figures, 1 table. Rejected at ICLR

2. CoupledFlow: One-Step Neural Operators for Coupled Multi-Physics PDEs

探讨了CoupledFlow: One-Step Operators Coupled Multi-Physics PDEs的相关问题,值得关注。

来源:Accepted NeurIPS 2026

3. Convergence analysis of physics-informed neural networks for subdiffusion

对Convergence analysis physics-informed networks subdiffusion进行了系统性理论分析,给出了收敛性及误差界。

4. Parameterized Physics-Informed Neural Networks for Efficient Simulation of Embedded Microfluidic Chip Cooling

探索了Parameterized Physics-Informed Networks Simulation Embedded Microfluidic在实际场景中的应用效果。

5. Rethinking How to Leverage Cheap Physics for Data-Efficient Neural Operators

探讨了Rethinking Leverage Cheap Physics Data-Efficient Operators的相关问题,值得关注。

6. A systematic comparison of physics-informed neural networks and mechanistic models for epidemic …

探讨了systematic comparison physics-informed networks mechanistic models的相关问题,值得关注。

7. Well-Balanced WE-PINNs for the Shallow Water Equations over Smooth and Discontinuous Bathymetries

探讨了Well-Balanced WE-PINNs Shallow Water Equations over的相关问题,值得关注。

8. Physics-informed neural networks in the small-data regime: A controlled comparison with classical …

探讨了Physics-informed networks small-data regime: controlled comparison的相关问题,值得关注。

9. On the Global Convergence of Stochastic Adam for Physics-Informed Neural Networks

对Global Convergence Stochastic Adam Physics-Informed Networks进行了系统性理论分析,给出了收敛性及误差界。

10. ALETHEIA: A Multi-Frequency Eddy Current Pulsed Thermography Dataset for Neural Operator Learning …

提出一种ALETHEIA: Multi-Frequency Eddy Current Pulsed Thermography的方法,在多个任务上验证了其有效性。

来源:62 pages, 19 figures


数据来自 arXiv API(scripts/fetch-arxiv-digest.mjs),仅收录元数据与站点自撰推荐理由,不转载摘要全文。 收录标准:目标分类(cs.LG / physics.flu-dyn / physics.comp-ph / math.NA)× 核心词相关度 ≥ 2 × 最近 45 天提交 × 与站内已收录论文不重复。