Incompressible fluid flow around a cylinder is one of the classical problems in fluid-dynamics with strong relevance with many real-world engineering problems, for example, design of offshore structures or design of a pin-fin heat exchanger. Thus learning a high-accuracy surrogate for this problem can demonstrate the efficacy of a novel machine learning approach. In this work, we propose a physics-informed neural network (PINN) architecture for learning the relationship between simulation output and the underlying geometry and boundary conditions. In addition to using a physics-based regularization term, the proposed approach also exploits the underlying physics to learn a set of Fourier features, i.e. frequency and phase offset parameters, and then use them for predicting flow velocity and pressure over the spatio-temporal domain. We demonstrate this approach by predicting simulation results over out of range time interval and for novel design conditions. Our results show that incorporation of Fourier features improves the generalization performance over both temporal domain and design space.
Learning Fluid-Structure Interaction with Physics-Informed Machine Learning and Immersed Boundary Methods
Computational Science and Engineering Department, Informatics Institute Istanbul Technical University（信息学院伊斯坦布尔技术大学计算科学与工程系）；Scientific Computing, Rutherford Appleton Laboratory Science and Technology Facilities Council (STFC)（科学与技术设施理事会（STFC）科学计算部）
PINNs for the Solution of the Hyperbolic Buckley-Leverett Problem with a Non-convex Flux Function
Khalifa University of Science and Technology（哈利法科技大学）
两种不混溶流体的驱替是多孔介质中流体流动的一个普遍问题。这样的问题可以作为偏微分方程（PDE）提出，通常称为Buckley-Leverett（B-L）问题。B-L问题是一个非线性双曲守恒定律，众所周知，使用传统的数值方法很难求解。在这里，我们使用物理信息神经网络（PINNs）解决具有非凸通量函数的前向双曲B-L问题。本文的贡献是双重的。首先，我们提出了一个PINN方法来解决双曲B-L问题嵌入Oleinik熵条件到神经网络的残差。我们在剩余损失中不使用扩散项（人工粘性），而是依赖于偏微分方程的强形式。其次，我们使用Adam优化器和基于残差的自适应细化（RAR）算法来实现超低损失而无需加权。我们的解决方案可以准确地捕捉冲击波阵面，并产生一个准确的整体解决方案。我们报告了2 x 10-2的L2验证误差和1 x 10-6的L2损失。所提出的方法不需要任何额外的正则化或加权损失，以获得这样的准确的解决方案。
Finite basis physics-informed neural networks with hard constraints for viscous fluid flow in highly perforated domains
Korea Advanced Institute of Science and Technology（韩国科学技术院）；Weierstrass-Institute for Applied Analysis and Stochastics（魏尔斯特拉斯应用分析与随机学研究所）；Leibniz-Institut für Verbundwerkstoffe GmbH（莱布尼茨复合材料有限公司研究所）
AQ-PINNs: Attention-Enhanced Quantum Physics-Informed Neural Networks for Carbon-Efficient Climate Modeling
SVKM’s Dwarkadas J. Sanghvi College of Engineering（SVKM德瓦卡达斯·J·桑格维工程学院）；New York University Abu Dhabi (NYUAD)（纽约大学阿布扎比分校）；NYUAD Research Institute（纽约大学阿布扎比分校研究院）；The Maersk Mc-Kinney Moller Institute（马士基麦克-金尼·莫勒研究所）；University of Southern Denmark（南丹麦大学）
Scale-PINN: Learning Efficient Physics-Informed Neural Networks Through Sequential Correction
Institute of High Performance Computing（高性能计算研究所）；Centre for Frontier AI Research（前沿人工智能研究中心）；School of Mechanical Engineering（机械工程学院）；College of Computing and Data Science（计算与数据科学学院）
From PINNs to PIKANs: Recent Advances in Physics-Informed Machine Learning
Brown University（布朗大学）；Division of Applied Mathematics, Brown University（布朗大学应用数学系）；School of Engineering, Brown University（布朗大学工程学院）；Center for Biomedical Engineering, Brown University（布朗大学生物医学工程中心）
Simulating Three-dimensional Turbulence with Physics-informed Neural Networks
Institution for Foundation of Data Science, Yale University（数据科学基础研究所，耶鲁大学）；Department of Mechanical Engineering and Applied Mechanics, University of Pennsylvania（机械工程与应用力学系，宾夕法尼亚大学）；Sibley School of Mechanical and Aerospace Engineering, Cornell University（机械与航空航天工程学院，康奈尔大学）；Advanced Computing, Mathematics and Data Division, Pacific Northwest National Laboratory（先进计算、数学和数据部门，太平洋西北国家实验室）
