Control of energy systems based on self-learning algorithms

Project Description

The worldwide expansion of decentralized generation units, such as photovoltaic systems and wind power plants, means that the generation of electrical energy depends on the fluctuating supply of renewable energy sources. In the project "Control of energy systems based on self-learning algorithms", new self-learning control algorithms for energy systems are investigated and compared with classical control algorithms. Models written in Python will make it possible to create several self-learning agents and simulate them in a common environment.

It is also planned to model further components such as electric charging stations in order to take into account the influence of electromobility on the power system.

Project Period

2017-05-16 –

Project Leaders

Marco Pruckner

Project Members

Niklas Ebell

Related Publications

• Ebell N., Pruckner M.:
  Coordinated Multi-Agent Reinforcement Learning for Swarm Battery Control
  2018 IEEE Canadian Conference on Electrical & Computer Engineering (CCECE) (Quebec Stadt, Quebec, 2018-05-21 - 2018-05-23)
  In: 2018 IEEE Canadian Conference on Electrical Computer Engineering (CCECE) 2018
  DOI: 10.1109/CCECE.2018.8447851
  BibTeX: Download
• Ebell N., Heinrich F., Schlund J., Pruckner M.:
  Reinforcement Learning Control Algorithm for a PV-Battery-System Providing Frequency Containment Reserve Power
  2018 IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids (SmartGridComm) (Aalborg, 2018-10-29 - 2018-10-31)
  In: 2018 IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids (SmartGridComm) 2018
  DOI: 10.1109/SmartGridComm.2018.8587480
  BibTeX: Download
• Ebell N., Gütlein M., Pruckner M.:
  Sharing of Energy Among Cooperative Households Using Distributed Multi-Agent Reinforcement Learning
  2019 IEEE Innovative Smart Grid Technologies Europe (Bucharest, 2019-09-29 - 2019-10-02)
  DOI: 10.1109/ISGTEurope.2019.8905520
  URL: https://ieeexplore.ieee.org/document/8905520
  BibTeX: Download
• Tuchnitz F., Ebell N., Schlund J., Pruckner M.:
  Development and Evaluation of a Smart Charging Strategy for an Electric Vehicle Fleet Based on Reinforcement Learning
  In: Applied Energy 285 (2021)
  ISSN: 0306-2619
  DOI: 10.1016/j.apenergy.2020.116382
  BibTeX: Download
• Ebell N., Pruckner M.:
  Benchmarking a Decentralized Reinforcement Learning Control Strategy for an Energy Community
  2021 IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids (SmartGridComm) (Aachen, 2021-10-25 - 2021-10-28)
  DOI: 10.1109/smartgridcomm51999.2021.9632323
  BibTeX: Download