KIT | KIT-Bibliothek | Impressum | Datenschutz

PowerSUME

Weiskopf, Thorsten ORCID iD icon 1; Miskiw, Kim K. ORCID iD icon 2; Treml, Niklas; Stelzer, Jonathan ORCID iD icon 1
1 Institut für Industriebetriebslehre und Industrielle Produktion (IIP), Karlsruher Institut für Technologie (KIT)
2 Institut für Wirtschaftsinformatik und Marketing (IISM), Karlsruher Institut für Technologie (KIT)

Abstract (englisch):

PowerSUME serves as an interface code that makes the reinforcement‑learning capabilities of the ASSUME framework usable for other agent‑based electricity‑market models. It enables data exchange via an HTTP interface, allowing the electricity‑market model to transmit power‑plant data and forecast information to PowerSUME. PowerSUME then manages the learning environment and, using ASSUME, returns bids.

For more details about the ASSUME framework and its capabilities, please consult the project website https://assume-project.de/. Note that we currently rely on a specific commit of ASSUME (see Prerequisites) to ensure compatibility; the latest version may require adjustments.

At present, only a single price can be passed per unit, and it always markets the maximum available generation capacity.

To use the interface effectively, an additional market model is required that can communicate with the interface according to the endpoints and data contract shown below. In general we only recommend PowerSUME in case, your own model brings an advanadge over ASSUME's market simulation capabilities, for example because you have an existing model with specific features or a preferred modeling approach. ... mehr


Download
Originalveröffentlichung
DOI: 10.5281/zenodo.19125583
Zugehörige Institution(en) am KIT Institut für Industriebetriebslehre und Industrielle Produktion (IIP)
Institut für Wirtschaftsinformatik (WIN)
Publikationstyp Forschungsdaten
Publikationsdatum 20.03.2026
Identifikator KITopen-ID: 1000192792
Lizenz GNU Affero General Public License v3.0 only
Projektinformation BETS (BMWE, 03EI1069B)
Art der Forschungsdaten Software
Nachgewiesen in OpenAlex
KIT – Die Universität in der Helmholtz-Gemeinschaft
KITopen Landing Page