Learning to run a power network challenge for training topology controllers
This addresses the need for more flexible and less costly grid operations for power network operators, though it appears incremental as it builds on prior work on branch switching.
The paper tackles the problem of power grid topology reconfiguration, a complex and combinatorial task not solvable by existing optimal power flow solvers, by proposing a framework using imitation and reinforcement learning, with results from a challenge showing performance improvements but not specifying concrete numbers.
For power grid operations, a large body of research focuses on using generation redispatching, load shedding or demand side management flexibilities. However, a less costly and potentially more flexible option would be grid topology reconfiguration, as already partially exploited by Coreso (European RSC) and RTE (French TSO) operations. Beyond previous work on branch switching, bus reconfigurations are a broader class of action and could provide some substantial benefits to route electricity and optimize the grid capacity to keep it within safety margins. Because of its non-linear and combinatorial nature, no existing optimal power flow solver can yet tackle this problem. We here propose a new framework to learn topology controllers through imitation and reinforcement learning. We present the design and the results of the first "Learning to Run a Power Network" challenge released with this framework. We finally develop a method providing performance upper-bounds (oracle), which highlights remaining unsolved challenges and suggests future directions of improvement.