Real-Time Bidding Strategies from Micro-Grids Using Reinforcement Learning

Thumbnail Image
Files
CIRED 2018 Ljubljana WS - 0440 - 21014.pdf (233.72 KB)
Paper number
0440
Working Group Number
Conference name
CIRED 2018 Ljubljana Workshop
Conference date
7 - 8 June 2018
Conference location
Ljubljana, Slovenia
Peer-reviewed
Yes
Short title
Convener
Authors
Boukas, Ioannis, University of Liège, Belgium
Ernst, Damien, University of Liège, Belgium
Cornélusse, Bertrand, University of Liège, Belgium
Abstract
We address the problem faced by the operator of a microgrid participating in a continuous real-time market. Themicrogrid consists of distributed generation, flexible loadsand a storage device. The goal of the microgrid operatoris the maximization of the profits over the entire tradinghorizon, while taking into account operational constraints.The variability of the Renewable Energy Sources (RES) isconsidered and the energy trading is modeled as a MarkovDecision Process. The problem is solved using reinforcement learning (RL). The resulting optimal real time bidding strategy of a microgrid is discussed.
Table of content
Keywords
Publisher
AIM
Date
2018-06-07
Published in
CIRED 2018 Ljubljana Workshop on Microgrids and Local Energy Communities
Permanent link to this record
https://www.cired-repository.org/handle/20.500.12455/1129
http://dx.doi.org/10.34890/163
ISSN
2032-9628
ISBN
978-2-9602415-1-8