Branching Dueling Q-Network Based Online Scheduling of a Microgrid With Distributed Energy Storage Systems

TL;DR

This paper proposes a deep reinforcement learning approach using Branching Dueling Q-Networks for online microgrid scheduling with multiple distributed energy storage systems, effectively managing high-dimensional decision spaces.

Contribution

It introduces a scalable DRL-based online optimization method that handles multiple BESSs without suffering from the curse of dimensionality.

Findings

The method achieves linear scalability with the number of BESSs.

Numerical simulations confirm the effectiveness of the proposed approach.

The approach improves microgrid operation under uncertainties.

Abstract

This letter investigates a Branching Dueling Q-Network (BDQ) based online operation strategy for a microgrid with distributed battery energy storage systems (BESSs) operating under uncertainties. The developed deep reinforcement learning (DRL) based microgrid online optimization strategy can achieve a linear increase in the number of neural network outputs with the number of distributed BESSs, which overcomes the curse of dimensionality caused by the charge and discharge decisions of multiple BESSs. Numerical simulations validate the effectiveness of the proposed method.