Get Started
Home
Topics
Search
Library
Research questionHow should graph representations be chosen for GNN-based reinforcement learning in power-grid topology control?Power-grid topology control depends on how network structure is encoded for the graph neural network. Physical, electrical-sensitivity, and hybrid representations may provide different levels of useful structure for reinforcement-learning agents.
AI
Evaluation & Benchmarks
Machine Learning
Reinforcement Learning
Latest papersRecent research connected to this question, newest first.A Comparative Study of Graph Representations for GNN-Based Power Grid Control in L2RPNThe evidence concerns controlled comparisons in the Learning to Run a Power Network (L2RPN) environment, covering physical-topology, electrical-sensitivity, and hybrid graph representations for GNN-based topology control. It supports conclusions about representation effects in this setting, not necessarily other grid-control environments or architectures.research paper · Sep 2, 2026
Related questions
How can graph learning leverage visual graph depictions for structural reasoning?How can power-grid graph classifiers predict cascading failures using interpretable, lightweight representations of operational edge states?How can representation-learning methods for dynamic heterogeneous graphs be compared across temporal granularities?How can graph neural networks capture long-range interactions without oversmoothing or oversquashing on diverse large graphs?