Wind Power Prediction using Multi-Task Gaussian Process Regression with Lagged Inputs

Francisco Javier Jara Avila; Timothy Verstraeten; Konstantinos Vratsinis; Ann Nowe; Jan Helsen

doi:10.1088/1742-6596/2505/1/012035

Wind Power Prediction using Multi-Task Gaussian Process Regression with Lagged Inputs

Francisco Javier Jara Avila, Timothy Verstraeten, Konstantinos Vratsinis, Ann Nowe, Jan Helsen

Onderzoeksoutput: Conference paper

3 Citaten (Scopus)

Samenvatting

Wind is a renewable energy source that has become more important in recent years. Wind turbines are equipped with a SCADA system, which allows for remote supervision of the wind farm. SCADA systems are customarily used to provide data averaged every 10 minutes. Nevertheless, recent literature suggests that more insights could be extracted with a higher granularity of data. In this work, a naive methodology based on Multi-Task Gaussian Process Regression is presented, in order to show how spatiotemporal modeling benefits power estimation. Using sparsity properties a model for possible power prediction is proposed. The model proposed performs better than the power curves provided by the manufacturer.

Originele taal-2	English
Titel	Journal of Physics: Conference Series
Plaats van productie	Visby, Sweden
Uitgeverij	IOP Publishing
Aantal pagina's	12
Volume	2505
Uitgave	1
DOI's	https://doi.org/10.1088/1742-6596/2505/1/012035
Status	Published - 22 jun. 2023

Publicatie series

Naam	Journal of Physics: Conference Series
ISSN van geprinte versie	1742-6588

Bibliografische nota

Funding Information:
This research was funded in the context of the Energy Transition Fund project Poseidon. The authors acknowledge VLAIO/Blue Cluster Supersized 4.0 ICON project (HBC.2019.0135), ICON Rainbow project, and SEAFD project. The research was supported by the Flemish Government under the “Onderzoeksprogramma Artificiële Intelligentie (AI) Vlaanderen” program. The authors are also grateful to the VSC Supercomputing Flanders centre for the support in the context of the VSC Cloud program.

Publisher Copyright:
© Published under licence by IOP Publishing Ltd.

Copyright:
Copyright 2023 Elsevier B.V., All rights reserved.

Toegang tot document

10.1088/1742-6596/2505/1/012035

Andere bestanden en links

Link naar publicatie in Scopus

Citeer dit

@inproceedings{58a24d596e6544d183070ef1a417c6e3,

title = "Wind Power Prediction using Multi-Task Gaussian Process Regression with Lagged Inputs",

abstract = "Wind is a renewable energy source that has become more important in recent years. Wind turbines are equipped with a SCADA system, which allows for remote supervision of the wind farm. SCADA systems are customarily used to provide data averaged every 10 minutes. Nevertheless, recent literature suggests that more insights could be extracted with a higher granularity of data. In this work, a naive methodology based on Multi-Task Gaussian Process Regression is presented, in order to show how spatiotemporal modeling benefits power estimation. Using sparsity properties a model for possible power prediction is proposed. The model proposed performs better than the power curves provided by the manufacturer.",

keywords = "Machine Learning, Wind Energy, Gaussian Process",

author = "{Jara Avila}, {Francisco Javier} and Timothy Verstraeten and Konstantinos Vratsinis and Ann Nowe and Jan Helsen",

note = "Funding Information: This research was funded in the context of the Energy Transition Fund project Poseidon. The authors acknowledge VLAIO/Blue Cluster Supersized 4.0 ICON project (HBC.2019.0135), ICON Rainbow project, and SEAFD project. The research was supported by the Flemish Government under the “Onderzoeksprogramma Artifici{\"e}le Intelligentie (AI) Vlaanderen” program. The authors are also grateful to the VSC Supercomputing Flanders centre for the support in the context of the VSC Cloud program. Publisher Copyright: {\textcopyright} Published under licence by IOP Publishing Ltd. Copyright: Copyright 2023 Elsevier B.V., All rights reserved.",

year = "2023",

month = jun,

day = "22",

doi = "10.1088/1742-6596/2505/1/012035",

language = "English",

volume = "2505",

series = "Journal of Physics: Conference Series",

publisher = "IOP Publishing",

booktitle = "Journal of Physics: Conference Series",

edition = "1",

}

TY - GEN

T1 - Wind Power Prediction using Multi-Task Gaussian Process Regression with Lagged Inputs

AU - Jara Avila, Francisco Javier

AU - Verstraeten, Timothy

AU - Vratsinis, Konstantinos

AU - Nowe, Ann

AU - Helsen, Jan

N1 - Funding Information: This research was funded in the context of the Energy Transition Fund project Poseidon. The authors acknowledge VLAIO/Blue Cluster Supersized 4.0 ICON project (HBC.2019.0135), ICON Rainbow project, and SEAFD project. The research was supported by the Flemish Government under the “Onderzoeksprogramma Artificiële Intelligentie (AI) Vlaanderen” program. The authors are also grateful to the VSC Supercomputing Flanders centre for the support in the context of the VSC Cloud program. Publisher Copyright: © Published under licence by IOP Publishing Ltd. Copyright: Copyright 2023 Elsevier B.V., All rights reserved.

PY - 2023/6/22

Y1 - 2023/6/22

N2 - Wind is a renewable energy source that has become more important in recent years. Wind turbines are equipped with a SCADA system, which allows for remote supervision of the wind farm. SCADA systems are customarily used to provide data averaged every 10 minutes. Nevertheless, recent literature suggests that more insights could be extracted with a higher granularity of data. In this work, a naive methodology based on Multi-Task Gaussian Process Regression is presented, in order to show how spatiotemporal modeling benefits power estimation. Using sparsity properties a model for possible power prediction is proposed. The model proposed performs better than the power curves provided by the manufacturer.

AB - Wind is a renewable energy source that has become more important in recent years. Wind turbines are equipped with a SCADA system, which allows for remote supervision of the wind farm. SCADA systems are customarily used to provide data averaged every 10 minutes. Nevertheless, recent literature suggests that more insights could be extracted with a higher granularity of data. In this work, a naive methodology based on Multi-Task Gaussian Process Regression is presented, in order to show how spatiotemporal modeling benefits power estimation. Using sparsity properties a model for possible power prediction is proposed. The model proposed performs better than the power curves provided by the manufacturer.

KW - Machine Learning

KW - Wind Energy

KW - Gaussian Process

UR - http://www.scopus.com/inward/record.url?scp=85163507202&partnerID=8YFLogxK

U2 - 10.1088/1742-6596/2505/1/012035

DO - 10.1088/1742-6596/2505/1/012035

M3 - Conference paper

VL - 2505

T3 - Journal of Physics: Conference Series

BT - Journal of Physics: Conference Series

PB - IOP Publishing

CY - Visby, Sweden

ER -

Wind Power Prediction using Multi-Task Gaussian Process Regression with Lagged Inputs

Samenvatting

Publicatie series

Bibliografische nota

Toegang tot document

Andere bestanden en links

Vingerafdruk

Citeer dit