File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1109/JSYST.2021.3114445
- Scopus: eid_2-s2.0-85118258322
- WOS: WOS:000732321900001
- Find via
Supplementary
- Citations:
- Appears in Collections:
Article: Forecast-Driven Stochastic Scheduling of a Virtual Power Plant in Energy and Reserve Markets
Title | Forecast-Driven Stochastic Scheduling of a Virtual Power Plant in Energy and Reserve Markets |
---|---|
Authors | |
Keywords | Bidirectional long short-term memory (BLSTM) Machine learning Renewable generation Virtual power plant (VPP) |
Issue Date | 2021 |
Citation | IEEE Systems Journal, 2021 How to Cite? |
Abstract | Virtual power plants (VPPs) offer a cost-effective solution to incentivize coordination between different resources participating in joint energy and reserve markets. However, emerging technologies such as storage and demand response cannot deliver flexibility over long periods due to inherent energy limitations. In this article, we, therefore, inform the day-ahead scheduling of VPPs with forecast scenarios of the balancing stage, which are complemented with information on nonshiftable load, renewable generation, and electricity prices. These multivariate scenarios (incorporating both time and cross-variable dependencies) are obtained using a machine learning framework in which probabilistic forecasts are converted into time trajectories using a copula-based sampling. The model is enriched with a detailed representation of the intraday decision stage wherein all flexible resources are dynamically allocated over the daily horizon. To ensure the reliability of the solution, we take into full consideration the revenues and unit-specific costs related to the activation of reserves. Outcomes show that improving the representation of the intraday dispatch stage while relying on representative balancing uncertainties are two complementary components for increasing the quality of the VPP day-ahead strategy, which ultimately fosters its economic value. |
Persistent Identifier | http://hdl.handle.net/10722/308900 |
ISSN | 2023 Impact Factor: 4.0 2023 SCImago Journal Rankings: 1.402 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Toubeau, Jean Franeois | - |
dc.contributor.author | Nguyen, Thuy Hai | - |
dc.contributor.author | Khaloie, Hooman | - |
dc.contributor.author | Wang, Yi | - |
dc.contributor.author | Vallee, Franeois | - |
dc.date.accessioned | 2021-12-08T07:50:22Z | - |
dc.date.available | 2021-12-08T07:50:22Z | - |
dc.date.issued | 2021 | - |
dc.identifier.citation | IEEE Systems Journal, 2021 | - |
dc.identifier.issn | 1932-8184 | - |
dc.identifier.uri | http://hdl.handle.net/10722/308900 | - |
dc.description.abstract | Virtual power plants (VPPs) offer a cost-effective solution to incentivize coordination between different resources participating in joint energy and reserve markets. However, emerging technologies such as storage and demand response cannot deliver flexibility over long periods due to inherent energy limitations. In this article, we, therefore, inform the day-ahead scheduling of VPPs with forecast scenarios of the balancing stage, which are complemented with information on nonshiftable load, renewable generation, and electricity prices. These multivariate scenarios (incorporating both time and cross-variable dependencies) are obtained using a machine learning framework in which probabilistic forecasts are converted into time trajectories using a copula-based sampling. The model is enriched with a detailed representation of the intraday decision stage wherein all flexible resources are dynamically allocated over the daily horizon. To ensure the reliability of the solution, we take into full consideration the revenues and unit-specific costs related to the activation of reserves. Outcomes show that improving the representation of the intraday dispatch stage while relying on representative balancing uncertainties are two complementary components for increasing the quality of the VPP day-ahead strategy, which ultimately fosters its economic value. | - |
dc.language | eng | - |
dc.relation.ispartof | IEEE Systems Journal | - |
dc.subject | Bidirectional long short-term memory (BLSTM) | - |
dc.subject | Machine learning | - |
dc.subject | Renewable generation | - |
dc.subject | Virtual power plant (VPP) | - |
dc.title | Forecast-Driven Stochastic Scheduling of a Virtual Power Plant in Energy and Reserve Markets | - |
dc.type | Article | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1109/JSYST.2021.3114445 | - |
dc.identifier.scopus | eid_2-s2.0-85118258322 | - |
dc.identifier.eissn | 1937-9234 | - |
dc.identifier.isi | WOS:000732321900001 | - |