.png){kind=icon}
.png){kind=icon}
Return
Towards generalization of graph neural networks for AC optimal power flow
.png){kind=icon}
OA
AI
O
J
DOI：10.1016/j.egyai.2026.100842.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• This paper presents HH-MPNN for approximating ACOPF solutions efficiently. • HH-MPNN combines heterogeneous GNN and a scalar transformer using physics- informed positional encoding. • Generalizes to N-1 contingencies; pretraining on smaller grids improves larger grids. • Evaluated on diverse ACOPF benchmarks, including PGLearn and GridFM-Datakit datasets.
Keywords:
Graph neural network
Generalization
Machine learning
Optimal power flow
Topology
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.