Physics-Informed GNN Coupled with ESN for Solving Forward Problems of Spatiotemporal Partial Differential Equations
Bibliographic record
Abstract
Partial Differential Equations (PDEs) are the foundation of modeling and simulation in numerous scientific and engineering fields. In recent years, breakthrough advancements in deep learning, particularly the rise of Physics-Informed Neural Networks (PINNs), have opened up a data-driven new paradigm for PDE solving and demonstrated enormous potential. However, PINN is essentially a global fitting method based on fully connected networks, and its core drawback is that the global fitting characteristics lead to a large amount of redundancy in high-order derivative calculations and insufficient modeling of spatiotemporal correlations. To address this, we propose the Physics-Informed E-GNN method, which modeling spatiotemporal features separately under a discrete learning framework to improve the accuracy of spatiotemporal prediction. Our method first discretizes the initial values of the PDE into a graph structure as input, feeds it into a Graph Neural Networks (GNN) to update the spatial features, and then inputs the updated feature vectors into an Echo State Networks (ESN) autoregressive module in the form of a time series to capture sequence correlations. We conducted comparative experiments on two classic partial differential equations (the 2D Burgers' equation and the 2D Convection-Diffusion equation) in irregular domains. The experimental results show that our proposed method achieves significant improvements in both solution accuracy and generality, and can effectively capture the complex patterns of changes in the PDE system.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".