Code and models for the paper:
Learning Topology Actions for Power Grid Control: A Graph-Based Soft-Label Imitation Learning Approach
Accepted at ECML PKDD 2025
Mohamed Hassouna, Clara Holzhüter, Malte Lehna, Matthijs de Jong, Jan Viebahn, Bernhard Sick, Christoph Scholz
[Pre-print Paper PDF]
The code and models will be released here.
This repository contains the official implementation of the SoftGNN agent, a novel Graph Neural Network (GNN)-based imitation learning framework for power grid topology control. The approach improves over traditional hard-label imitation learning by learning from soft labels that capture multiple viable actions for grid congestion mitigation. The agent operates in the Grid2Op L2RPN WCCI 2022 environment, outperforming both the expert and state-of-the-art RL agents.
- Soft-Label Generation: Learn from a distribution over viable actions rather than a single expert action.
- GNN-Based Architecture: Use Graph Attention Networks (GAT) to encode power grid topology.
- Action Feasibility Enhancements: Improved substation reconfiguration support and line-disconnection handling.
- N-1 Security Evaluation: Post-hoc contingency-aware action selection for increased robustness.
- Benchmarking: Evaluation against greedy expert, and SOTA DRL agents in the L2RPN WCCI 2022 environment.
For questions or collaborations, please contact: [email protected]