A Physics-Augmented GraphGPS Framework for the Reconstruction of 3D Riemann Problems from Sparse Data

2025/05/27 by Rami Cassia, Cassia, Rami, Rich R. Kerswell +1
Earth and Planetary Sciences · Engineering · Medicine · #3D Shape Modeling and Analysis #FOS: Computer and information sciences #FOS: Physical sciences #Fluid Dynamics (physics.flu-dyn) #Geological Modeling and Analysis #Machine Learning (cs.LG) #Medical Imaging Techniques and Applications

paper · pdf · doi:10.48550/arxiv.2505.21421

openalex publication_date 2025/05/27 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

In compressible fluid flow, reconstructing shocks, discontinuities, rarefactions, and their interactions from sparse measurements is an important inverse problem with practical applications. Moreover, physics-informed machine learning has recently become an increasingly popular approach for performing reconstructions tasks. In this work we explore a machine learning recipe, known as GraphGPS, for reconstructing canonical compressible flows known as 3D Riemann problems from sparse observations, in a physics-informed manner. The GraphGPS framework combines the benefits of positional encodings, local message-passing of graphs, and global contextual awareness, and we explore the latter two components through an ablation study. Furthermore, we modify the aggregation step of message-passing such that it is aware of shocks and discontinuities, resulting in sharper reconstructions of these features. Additionally, we modify message-passing such that information flows strictly from known nodes only, which results in computational savings, better training convergence, and no degradation of reconstruction accuracy. We also show that the GraphGPS framework outperforms numerous machine learning benchmarks.

Citations

• Deep Operator Learning for High-Fidelity Fluid Flow Field Reconstruction from Sparse Sensor Measurements
• FLRNet: A Deep Learning Method for Regressive Reconstruction of Flow Field From Limited Sensor Measurements
• Flow reconstruction in time-varying geometries using graph neural networks
• Godunov Loss Functions for Modelling of Hyperbolic Conservation Laws
• Graph Mamba: Towards Learning on Graphs with State Space Models
• Graph-Mamba: Towards Long-Range Graph Sequence Modeling with Selective State Spaces
• Gappy AE: A Nonlinear Approach for Gappy Data Reconstruction using Auto-Encoder
• Mamba: Linear-Time Sequence Modeling with Selective State Spaces
• Exphormer: Sparse Transformers for Graphs
• Graph Neural Networks for Aerodynamic Flow Reconstruction from Sparse Sensing
• Recipe for a General, Powerful, Scalable Graph Transformer
• On the Unreasonable Effectiveness of Feature propagation in Learning on Graphs with Missing Node Features
• Rethinking Graph Transformers with Spectral Attention
• A Generalization of Transformer Networks to Graphs
• An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
• Rethinking Attention with Performers
• Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention
• How Powerful are Graph Neural Networks?
• Graph Attention Networks
• Attention Is All You Need
• Inductive Representation Learning on Large Graphs
• Semi-Supervised Classification with Graph Convolutional Networks
• Effective Approaches to Attention-based Neural Machine Translation
• Neural Machine Translation by Jointly Learning to Align and Translate
• BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
• The Graph Neural Network Model
• Aerodynamic Data Reconstruction and Inverse Design Using Proper Orthogonal Decomposition

Related

• An Efficient Graph-Transformer Operator for Learning Physical Dynamics with Manifolds Embedding
• Graph Transformers for inverse physics: reconstructing flows around arbitrary 2D airfoils
• Scientific Computing Algorithms to Learn Enhanced Scalable Surrogates for Mesh Physics
• Simulating Liquids with Graph Networks
• Flow reconstruction in time-varying geometries using graph neural networks