Spatiotemporal attention networks for wind power forecasting

Xingbo Fu; Feng Gao; Jiang Wu; Xinyu Wei; Fangwei Duan

doi:10.1109/ICDMW.2019.00032

Spatiotemporal attention networks for wind power forecasting

Xingbo Fu
, Feng Gao
, Jiang Wu
, Xinyu Wei
, Fangwei Duan

Faculty of Electronic and Information Engineering

Xi'an Jiaotong University
Rutgers - The State University of New Jersey, New Brunswick
State Grid Corporation of China

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

42 Scopus citations

Abstract

Wind power is one of the most important renewable energy sources and accurate wind power forecasting is very significant for reliable and economic power system operation and control strategies. This paper proposes a novel framework with spatiotemporal attention networks (STAN) for wind power forecasting. This model captures spatial correlations among wind farms and temporal dependencies of wind power time series. First of all, we employ a multi-head self-attention mechanism to extract spatial correlations among wind farms. Then, temporal dependencies are captured by the Sequence-to-Sequence (Seq2Seq) model with a global attention mechanism. Finally, experimental results demonstrate that our model achieves better performance than other baseline approaches. Our work provides useful insights to capture non-Euclidean spatial correlations.

Original language	English
Title of host publication	Proceedings - 19th IEEE International Conference on Data Mining Workshops, ICDMW 2019
Editors	Panagiotis Papapetrou, Xueqi Cheng, Qing He
Publisher	IEEE Computer Society
Pages	149-154
Number of pages	6
ISBN (Electronic)	9781728146034
DOIs	https://doi.org/10.1109/ICDMW.2019.00032
State	Published - Nov 2019
Event	19th IEEE International Conference on Data Mining Workshops, ICDMW 2019 - Beijing, China Duration: 8 Nov 2019 → 11 Nov 2019

Publication series

Name	IEEE International Conference on Data Mining Workshops, ICDMW
Volume	2019-November
ISSN (Print)	2375-9232
ISSN (Electronic)	2375-9259

Conference

Conference	19th IEEE International Conference on Data Mining Workshops, ICDMW 2019
Country/Territory	China
City	Beijing
Period	8/11/19 → 11/11/19

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Keywords

Attention mechanism
Spatiotemporal attention networks
Wind power forecasting

Access to Document

10.1109/ICDMW.2019.00032

Cite this

Fu, X., Gao, F., Wu, J., Wei, X., & Duan, F. (2019). Spatiotemporal attention networks for wind power forecasting. In P. Papapetrou, X. Cheng, & Q. He (Eds.), Proceedings - 19th IEEE International Conference on Data Mining Workshops, ICDMW 2019 (pp. 149-154). Article 8955569 (IEEE International Conference on Data Mining Workshops, ICDMW; Vol. 2019-November). IEEE Computer Society. https://doi.org/10.1109/ICDMW.2019.00032

@inproceedings{9dc3abd6d859445393b1cff1824d01aa,

title = "Spatiotemporal attention networks for wind power forecasting",

abstract = "Wind power is one of the most important renewable energy sources and accurate wind power forecasting is very significant for reliable and economic power system operation and control strategies. This paper proposes a novel framework with spatiotemporal attention networks (STAN) for wind power forecasting. This model captures spatial correlations among wind farms and temporal dependencies of wind power time series. First of all, we employ a multi-head self-attention mechanism to extract spatial correlations among wind farms. Then, temporal dependencies are captured by the Sequence-to-Sequence (Seq2Seq) model with a global attention mechanism. Finally, experimental results demonstrate that our model achieves better performance than other baseline approaches. Our work provides useful insights to capture non-Euclidean spatial correlations.",

keywords = "Attention mechanism, Spatiotemporal attention networks, Wind power forecasting",

author = "Xingbo Fu and Feng Gao and Jiang Wu and Xinyu Wei and Fangwei Duan",

note = "Publisher Copyright: {\textcopyright} 2019 IEEE.; 19th IEEE International Conference on Data Mining Workshops, ICDMW 2019 ; Conference date: 08-11-2019 Through 11-11-2019",

year = "2019",

month = nov,

doi = "10.1109/ICDMW.2019.00032",

language = "英语",

series = "IEEE International Conference on Data Mining Workshops, ICDMW",

publisher = "IEEE Computer Society",

pages = "149--154",

editor = "Panagiotis Papapetrou and Xueqi Cheng and Qing He",

booktitle = "Proceedings - 19th IEEE International Conference on Data Mining Workshops, ICDMW 2019",

}

Fu, X, Gao, F, Wu, J, Wei, X & Duan, F 2019, Spatiotemporal attention networks for wind power forecasting. in P Papapetrou, X Cheng & Q He (eds), Proceedings - 19th IEEE International Conference on Data Mining Workshops, ICDMW 2019., 8955569, IEEE International Conference on Data Mining Workshops, ICDMW, vol. 2019-November, IEEE Computer Society, pp. 149-154, 19th IEEE International Conference on Data Mining Workshops, ICDMW 2019, Beijing, China, 8/11/19. https://doi.org/10.1109/ICDMW.2019.00032

Spatiotemporal attention networks for wind power forecasting. / Fu, Xingbo; Gao, Feng; Wu, Jiang et al.
Proceedings - 19th IEEE International Conference on Data Mining Workshops, ICDMW 2019. ed. / Panagiotis Papapetrou; Xueqi Cheng; Qing He. IEEE Computer Society, 2019. p. 149-154 8955569 (IEEE International Conference on Data Mining Workshops, ICDMW; Vol. 2019-November).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Spatiotemporal attention networks for wind power forecasting

AU - Fu, Xingbo

AU - Gao, Feng

AU - Wu, Jiang

AU - Wei, Xinyu

AU - Duan, Fangwei

PY - 2019/11

Y1 - 2019/11

N2 - Wind power is one of the most important renewable energy sources and accurate wind power forecasting is very significant for reliable and economic power system operation and control strategies. This paper proposes a novel framework with spatiotemporal attention networks (STAN) for wind power forecasting. This model captures spatial correlations among wind farms and temporal dependencies of wind power time series. First of all, we employ a multi-head self-attention mechanism to extract spatial correlations among wind farms. Then, temporal dependencies are captured by the Sequence-to-Sequence (Seq2Seq) model with a global attention mechanism. Finally, experimental results demonstrate that our model achieves better performance than other baseline approaches. Our work provides useful insights to capture non-Euclidean spatial correlations.

AB - Wind power is one of the most important renewable energy sources and accurate wind power forecasting is very significant for reliable and economic power system operation and control strategies. This paper proposes a novel framework with spatiotemporal attention networks (STAN) for wind power forecasting. This model captures spatial correlations among wind farms and temporal dependencies of wind power time series. First of all, we employ a multi-head self-attention mechanism to extract spatial correlations among wind farms. Then, temporal dependencies are captured by the Sequence-to-Sequence (Seq2Seq) model with a global attention mechanism. Finally, experimental results demonstrate that our model achieves better performance than other baseline approaches. Our work provides useful insights to capture non-Euclidean spatial correlations.

KW - Attention mechanism

KW - Spatiotemporal attention networks

KW - Wind power forecasting

UR - https://www.scopus.com/pages/publications/85078798535

U2 - 10.1109/ICDMW.2019.00032

DO - 10.1109/ICDMW.2019.00032

M3 - 会议稿件

AN - SCOPUS:85078798535

T3 - IEEE International Conference on Data Mining Workshops, ICDMW

SP - 149

EP - 154

BT - Proceedings - 19th IEEE International Conference on Data Mining Workshops, ICDMW 2019

A2 - Papapetrou, Panagiotis

A2 - Cheng, Xueqi

A2 - He, Qing

PB - IEEE Computer Society

T2 - 19th IEEE International Conference on Data Mining Workshops, ICDMW 2019

Y2 - 8 November 2019 through 11 November 2019

ER -

Spatiotemporal attention networks for wind power forecasting

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this