Coordinated Multi-Agent Reinforcement Learning for Swarm Battery Control

Ebell N, Pruckner M (2018)


Publication Language: English

Publication Type: Conference contribution, Conference Contribution

Publication year: 2018

Pages Range: 1-4

Conference Proceedings Title: 2018 IEEE Canadian Conference on Electrical Computer Engineering (CCECE)

Event location: Quebec Stadt, Quebec CA

ISBN: 978-1-5386-2410-4

DOI: 10.1109/CCECE.2018.8447851

Abstract

New ideas supporting the transition of the energy system are needed. To keep the electric grid in stable operation at times of high volatile supply from renewable energy sources one possibility is seen in distributed battery energy storage systems. They provide flexibility as a ancillary service for transmission system operators as well as improving self-sufficiency for residential buildings with photovoltaic systems. This research-in-progress paper presents an approach towards a coordinated multi-agent reinforcement learning-based swarm battery control. The goal is to use reinforcement learning to manage the battery’s power flows between the battery, a photovoltaic system, a household’s electric load and the electric grid. Our approach includes the use of the battery to offer frequency containment reserve as well as to improve energy self-sufficiency. As a last step, we compare the performance of our algorithm with a rule-based approach defining the same system configuration and objective.

Authors with CRIS profile

Related research project(s)

How to cite

APA:

Ebell, N., & Pruckner, M. (2018). Coordinated Multi-Agent Reinforcement Learning for Swarm Battery Control. In 2018 IEEE Canadian Conference on Electrical Computer Engineering (CCECE) (pp. 1-4). Quebec Stadt, Quebec, CA.

MLA:

Ebell, Niklas, and Marco Pruckner. "Coordinated Multi-Agent Reinforcement Learning for Swarm Battery Control." Proceedings of the 2018 IEEE Canadian Conference on Electrical & Computer Engineering (CCECE), Quebec Stadt, Quebec 2018. 1-4.

BibTeX: Download