Learning to run a power network challenge for training topology\n controllers

2019/12/05 by Antoine Marot, Benjamin Donnot, Marot, Antoine +11 · 1 citation
Engineering · #FOS: Computer and information sciences #FOS: Electrical engineering #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Microgrid Control and Optimization #Optimal Power Flow Distribution #Signal Processing (eess.SP) #Smart Grid Security and Resilience #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.1912.04211

openalex publication_date 2019/12/05 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

For power grid operations, a large body of research focuses on using\ngeneration redispatching, load shedding or demand side management\nflexibilities. However, a less costly and potentially more flexible option\nwould be grid topology reconfiguration, as already partially exploited by\nCoreso (European RSC) and RTE (French TSO) operations. Beyond previous work on\nbranch switching, bus reconfigurations are a broader class of action and could\nprovide some substantial benefits to route electricity and optimize the grid\ncapacity to keep it within safety margins. Because of its non-linear and\ncombinatorial nature, no existing optimal power flow solver can yet tackle this\nproblem. We here propose a new framework to learn topology controllers through\nimitation and reinforcement learning. We present the design and the results of\nthe first "Learning to Run a Power Network" challenge released with this\nframework. We finally develop a method providing performance upper-bounds\n(oracle), which highlights remaining unsolved challenges and suggests future\ndirections of improvement.\n

Citations

• Hierarchical Decision Making In Electricity Grid Management
• Safe and Nested Subgame Solving for Imperfect-Information Games
• Reinforcement Learning: An Introduction

Cited by

• Think Twice, Act Once: A Co-Evolution Framework of LLM and RL for Large-Scale Decision Making

Related

• Optimizing Power Grid Topologies with Reinforcement Learning: A Survey of Methods and Challenges
• Curriculum Based Reinforcement Learning of Grid Topology Controllers to Prevent Thermal Cascading
• Towards Efficient Multi-Objective Optimisation for Real-World Power Grid Topology Control
• Multi-Objective Reinforcement Learning for Power Grid Topology Control
• Reinforcement Learning for Electricity Network Operation