Multi Power-Market Bidding: Stochastic Optimization and Reinforcement Learning

Date
2024-01-03
Authors
Miskiw, Kim
Harder, Nick
Staudt, Philipp
Contributor
Advisor
Department
Instructor
Depositor
Speaker
Researcher
Consultant
Interviewer
Annotator
Journal Title
Journal ISSN
Volume Title
Publisher
Volume
Number/Issue
Starting Page
3062
Ending Page
Alternative Title
Abstract
The growing importance of short-term electricity trading over independent subsequent markets in Europe presents market participants with intricate decision challenges. Established solutions based on stochastic programs are often used but suffer from shortcomings such as the curse of dimensionality in multi-stage decision processes. Reinforcement learning is a promising alternative. However, best practices for the comparison of the two approaches and the ex-post evaluation of reinforcement learning are not yet established. In this paper, we offer a comparison of stochastic programs and reinforcement learning and propose measures for a comparative performance evaluation between the two approaches. We demonstrate them on an empirical case study over subsequent market stages of the German market zone within the coupled European power market.
Description
Keywords
Policy, Markets, and Analytics, multi-market bidding, reinforcement learning, sequential decision problems, stochastic optimisation
Citation
Extent
10 pages
Format
Geographic Location
Time Period
Related To
Proceedings of the 57th Hawaii International Conference on System Sciences
Table of Contents
Rights
Attribution-NonCommercial-NoDerivatives 4.0 International
Rights Holder
Local Contexts
Email libraryada-l@lists.hawaii.edu if you need this content in ADA-compliant format.