Data-augmented sequential deep learning for wind power forecasting
Accurate wind power forecasting plays a critical role in the operation of wind parks and the dispatch of wind energy into the power grid. With excellent automatic pattern recognition and nonlinear mapping ability for big data, deep learning is increasingly employed in wind power forecasting. However...
Published in: | Energy Conversion and Management |
---|---|
Main Authors: | , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Elsevier
2021
|
Subjects: | |
Online Access: | https://hdl.handle.net/10037/23515 https://doi.org/10.1016/j.enconman.2021.114790 |
id |
ftunivtroemsoe:oai:munin.uit.no:10037/23515 |
---|---|
record_format |
openpolar |
spelling |
ftunivtroemsoe:oai:munin.uit.no:10037/23515 2023-05-15T14:26:06+02:00 Data-augmented sequential deep learning for wind power forecasting Chen, Hao Birkelund, Yngve Qixia, Zhang 2021-11-15 https://hdl.handle.net/10037/23515 https://doi.org/10.1016/j.enconman.2021.114790 eng eng Elsevier Chen, H. (2022). Data-driven Arctic wind energy analysis by statistical and machine learning approaches. (Doctoral thesis). https://hdl.handle.net/10037/26938 Energy Conversion and Management Chen, Birkelund. Data-augmented sequential deep learning for wind power forecasting. Energy Conversion and Management. 2021 FRIDAID 1942745 doi:10.1016/j.enconman.2021.114790 0196-8904 1879-2227 https://hdl.handle.net/10037/23515 openAccess Copyright 2021 The Author(s) VDP::Mathematics and natural science: 400::Physics: 430 VDP::Matematikk og Naturvitenskap: 400::Fysikk: 430 VDP::Technology: 500 VDP::Teknologi: 500 Journal article Tidsskriftartikkel Peer reviewed publishedVersion 2021 ftunivtroemsoe https://doi.org/10.1016/j.enconman.2021.114790 2022-10-05T23:00:52Z Accurate wind power forecasting plays a critical role in the operation of wind parks and the dispatch of wind energy into the power grid. With excellent automatic pattern recognition and nonlinear mapping ability for big data, deep learning is increasingly employed in wind power forecasting. However, salient realities are that in-situ measured wind data are relatively expensive and inaccessible and correlation between steps is omitted in most multistep wind power forecasts. This paper is the first time that data augmentation is applied to wind power forecasting by systematically summarizing and proposing both physics-oriented and data-oriented time-series wind data augmentation approaches to considerably enlarge primary datasets, and develops deep encoder-decoder long short-term memory networks that enable sequential input and sequential output for wind power forecasting. The proposed augmentation techniques and forecasting algorithm are deployed on five turbines with diverse topographies in an Arctic wind park, and the outcomes are evaluated against benchmark models and different augmentations. The main findings reveal that on one side, the average improvement in RMSE of the proposed forecasting model over the benchmarks is 33.89%, 10.60%, 7.12%, and 4.27% before data augmentations, and increases to 40.63%, 17.67%, 11.74%, and 7.06%, respectively, after augmentations. The other side unveils that the effect of data augmentations on prediction is intricately varying, but for the proposed model with and without augmentations, all augmentation approaches boost the model outperformance from 7.87% to 13.36% in RMSE, 5.24% to 8.97% in MAE, and similarly over 12% in QR90. Finally, data-oriented augmentations, in general, are slightly better than physics-driven ones. Article in Journal/Newspaper Arctic Arctic University of Tromsø: Munin Open Research Archive Arctic Energy Conversion and Management 248 114790 |
institution |
Open Polar |
collection |
University of Tromsø: Munin Open Research Archive |
op_collection_id |
ftunivtroemsoe |
language |
English |
topic |
VDP::Mathematics and natural science: 400::Physics: 430 VDP::Matematikk og Naturvitenskap: 400::Fysikk: 430 VDP::Technology: 500 VDP::Teknologi: 500 |
spellingShingle |
VDP::Mathematics and natural science: 400::Physics: 430 VDP::Matematikk og Naturvitenskap: 400::Fysikk: 430 VDP::Technology: 500 VDP::Teknologi: 500 Chen, Hao Birkelund, Yngve Qixia, Zhang Data-augmented sequential deep learning for wind power forecasting |
topic_facet |
VDP::Mathematics and natural science: 400::Physics: 430 VDP::Matematikk og Naturvitenskap: 400::Fysikk: 430 VDP::Technology: 500 VDP::Teknologi: 500 |
description |
Accurate wind power forecasting plays a critical role in the operation of wind parks and the dispatch of wind energy into the power grid. With excellent automatic pattern recognition and nonlinear mapping ability for big data, deep learning is increasingly employed in wind power forecasting. However, salient realities are that in-situ measured wind data are relatively expensive and inaccessible and correlation between steps is omitted in most multistep wind power forecasts. This paper is the first time that data augmentation is applied to wind power forecasting by systematically summarizing and proposing both physics-oriented and data-oriented time-series wind data augmentation approaches to considerably enlarge primary datasets, and develops deep encoder-decoder long short-term memory networks that enable sequential input and sequential output for wind power forecasting. The proposed augmentation techniques and forecasting algorithm are deployed on five turbines with diverse topographies in an Arctic wind park, and the outcomes are evaluated against benchmark models and different augmentations. The main findings reveal that on one side, the average improvement in RMSE of the proposed forecasting model over the benchmarks is 33.89%, 10.60%, 7.12%, and 4.27% before data augmentations, and increases to 40.63%, 17.67%, 11.74%, and 7.06%, respectively, after augmentations. The other side unveils that the effect of data augmentations on prediction is intricately varying, but for the proposed model with and without augmentations, all augmentation approaches boost the model outperformance from 7.87% to 13.36% in RMSE, 5.24% to 8.97% in MAE, and similarly over 12% in QR90. Finally, data-oriented augmentations, in general, are slightly better than physics-driven ones. |
format |
Article in Journal/Newspaper |
author |
Chen, Hao Birkelund, Yngve Qixia, Zhang |
author_facet |
Chen, Hao Birkelund, Yngve Qixia, Zhang |
author_sort |
Chen, Hao |
title |
Data-augmented sequential deep learning for wind power forecasting |
title_short |
Data-augmented sequential deep learning for wind power forecasting |
title_full |
Data-augmented sequential deep learning for wind power forecasting |
title_fullStr |
Data-augmented sequential deep learning for wind power forecasting |
title_full_unstemmed |
Data-augmented sequential deep learning for wind power forecasting |
title_sort |
data-augmented sequential deep learning for wind power forecasting |
publisher |
Elsevier |
publishDate |
2021 |
url |
https://hdl.handle.net/10037/23515 https://doi.org/10.1016/j.enconman.2021.114790 |
geographic |
Arctic |
geographic_facet |
Arctic |
genre |
Arctic Arctic |
genre_facet |
Arctic Arctic |
op_relation |
Chen, H. (2022). Data-driven Arctic wind energy analysis by statistical and machine learning approaches. (Doctoral thesis). https://hdl.handle.net/10037/26938 Energy Conversion and Management Chen, Birkelund. Data-augmented sequential deep learning for wind power forecasting. Energy Conversion and Management. 2021 FRIDAID 1942745 doi:10.1016/j.enconman.2021.114790 0196-8904 1879-2227 https://hdl.handle.net/10037/23515 |
op_rights |
openAccess Copyright 2021 The Author(s) |
op_doi |
https://doi.org/10.1016/j.enconman.2021.114790 |
container_title |
Energy Conversion and Management |
container_volume |
248 |
container_start_page |
114790 |
_version_ |
1766298579738034176 |