Offline Reinforcement Learning for Microgrid Voltage Regulation

TL;DR

This paper explores offline reinforcement learning algorithms to regulate voltage in microgrids with solar power, enabling effective control without online interaction by training on pre-collected data, demonstrated on IEEE 33-bus system.

Contribution

It introduces the application of offline reinforcement learning to microgrid voltage regulation, addressing safety and technical constraints of online interaction.

Findings

Offline RL algorithms effectively regulate voltage in microgrids.

The approach works with low-quality offline datasets.

Demonstrated on IEEE 33-bus system.

Abstract

This paper presents a study on using different offline reinforcement learning algorithms for microgrid voltage regulation with solar power penetration. When environment interaction is unviable due to technical or safety reasons, the proposed approach can still obtain an applicable model through offline-style training on a previously collected dataset, lowering the negative impact of lacking online environment interactions. Experiment results on the IEEE 33-bus system demonstrate the feasibility and effectiveness of the proposed approach on different offline datasets, including the one with merely low-quality experience.