Learning to run a power network challenge for training topology controllers

TL;DR

This paper introduces a novel framework for training power grid topology controllers using imitation and reinforcement learning, aiming to optimize grid reconfiguration for safety and capacity.

Contribution

It presents the first challenge for learning power network topology control and develops an upper-bound method to evaluate potential improvements.

Findings

Successful training of topology controllers via imitation and reinforcement learning

Benchmark results demonstrating the effectiveness of the proposed framework

Identification of remaining challenges and future research directions

Abstract

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.