Reinforcement Learning Optimizes Power Dispatch in Decentralized Power Grid

TL;DR

This paper introduces GC-PPO, a reinforcement learning approach using graph convolutional networks to optimize power dispatch and improve frequency stability in decentralized power grids, demonstrated on the UK grid.

Contribution

The paper presents a novel GC-PPO method that effectively manages power dispatch in decentralized grids, outperforming classical control methods.

Findings

GC-PPO reduces frequency fluctuations more effectively than classical methods.

The approach enhances grid stability and reliability.

Demonstrated on the UK power grid with promising results.

Abstract

Effective frequency control in power grids has become increasingly important with the increasing demand for renewable energy sources. Here, we propose a novel strategy for resolving this challenge using graph convolutional proximal policy optimization (GC-PPO). The GC-PPO method can optimally determine how much power individual buses dispatch to reduce frequency fluctuations across a power grid. We demonstrate its efficacy in controlling disturbances by applying the GC-PPO to the power grid of the UK. The performance of GC-PPO is outstanding compared to the classical methods. This result highlights the promising role of GC-PPO in enhancing the stability and reliability of power systems by switching lines or decentralizing grid topology.