Power Prediction Method for Offshore Wind Farms Based on Temporal Convolutional Network-Transformer

WANG Xi; CHEN Xinyi

doi:10.12096/j.2096-4528.pgt.24251

English Version Journal Login

您当前的位置：

首页 >

文章列表页 >

Power Prediction Method for Offshore Wind Farms Based on Temporal Convolutional Network-Transformer

New Energy | 更新时间：2026-06-29

- Power Prediction Method for Offshore Wind Farms Based on Temporal Convolutional Network-Transformer
- Power Generation Technology Vol. 46, Issue 6, Pages: 1123-1132(2025)
- 作者机构：
  
  哈尔滨理工大学计算机科学与技术学院，黑龙江省哈尔滨市 150080
- 作者简介：
- 基金信息：
  
  National Key Research and Development Program of China(2023YFB4403500)
- DOI：10.12096/j.2096-4528.pgt.24251
  CLC： TK 89;TM 614
- Received：12 December 2024，
  
  Revised：2025-03-28，
  
  Published：31 December 2025
- 稿件说明：
移动端阅览
王曦,陈心怡.一种基于时序卷积网络-Transformer的海上风电功率预测方法[J].发电技术,2025,46(06):1123-1132.

WANG Xi,CHEN Xinyi.Power Prediction Method for Offshore Wind Farms Based on Temporal Convolutional Network-Transformer[J].Power Generation Technology,2025,46(06):1123-1132.
王曦,陈心怡.一种基于时序卷积网络-Transformer的海上风电功率预测方法[J].发电技术,2025,46(06):1123-1132. DOI： 10.12096/j.2096-4528.pgt.24251.

WANG Xi,CHEN Xinyi.Power Prediction Method for Offshore Wind Farms Based on Temporal Convolutional Network-Transformer[J].Power Generation Technology,2025,46(06):1123-1132. DOI： 10.12096/j.2096-4528.pgt.24251.

摘要

目的

海上风电对于实现碳达峰、碳中和目标具有重要意义，但其功率预测较为困难，为此提出了一种融合时序卷积网络(temporal convolutional network，TCN)与Transformer架构的海上风电功率预测新方法。

方法

首先，引入离散余弦变换(discrete cosine transform，DCT)技术，旨在有效提取时间序列数据的频域特征，充分挖掘频域特性来优化特征权重的分配。在此基础上，构建了TCN-Transformer复合模型，该模型能够充分整合时域与经过DCT处理的频域信息，实现对数据的深度学习与精准预测。

结果

选取某海上风电场的2组实际运行数据开展对比实验，结果表明，所提方法不仅在特征提取方面表现出色，还显著地提升了发电功率预测精度，充分验证了其有效性与应用价值。

结论

融合了DCT技术的TCN-Transformer模型在海上风电功率预测方面展现出了卓越的预测性能，相较于其他对比模型，其预测精度得到了显著提升，为海上风电场的高效运行与优化管理提供了有力支持。

Abstract

Objectives

Offshore wind power is of great significance for achieving carbon peaking and carbon neutrality goals

yet its power prediction remains challenging. To address this issue

this study proposes a novel offshore wind power prediction method integrating temporal convolutional network (TCN) with the Transformer framework.

Methods

Firstly

discrete cosine transform (DCT) technology is incorporated to effectively extract frequency-domain features from time-series data

optimizing feature weight allocation by fully exploiting frequency-domain characteristics. Based on this

a hybrid TCN-Transformer model is constructed

which is capable of comprehensively integrating both time-domain and DCT-processed frequency-domain information

enabling the deep learning and accurate prediction of the data.

Results

Comparative experiments are conducted using two sets of actual operational data from an offshore wind farm. The results show that the proposed method not only performs well in feature extraction but also significantly improves the accuracy of power prediction

fully verifying its effectiveness and application value.

Conclusions

The TCN-Transformer model integrated with DCT technology demonstrates exceptional prediction performance in offshore wind power prediction. Compared to other benchmark models

it achieves a significant improvement in prediction accuracy. This approach provides robust support for the efficient operation and optimal management of offshore wind farms.

关键词

Keywords

references

严新荣，张宁宁，马奎超，等．我国海上风电发展现状与趋势综述 [J ] ．发电技术， 2024 ， 45 ( 1 )： 1 - 12 ． doi: 10.12096/j.2096-4528.pgt.23093 http://dx.doi.org/10.12096/j.2096-4528.pgt.23093

YAN X R ， ZHANG N N ， MA K C ， et al ． Overview of current situation and trend of offshore wind power development in China [J ] ． Power Generation Technology ， 2024 ， 45 ( 1 )： 1 - 12 ． doi: 10.12096/j.2096-4528.pgt.23093 http://dx.doi.org/10.12096/j.2096-4528.pgt.23093

陈龙，刘璐洁，符杨，等．计及高风速和低风速影响的海上风电机组两阶段滚动优化维护策略 [J ] ．智慧电力， 2024 ， 52 ( 7 )： 80 - 87 ．

CHEN L ， LIU L J ， FU Y ， et al ． Two-stage rolling optimized maintenance strategy for offshore wind turbines considering the influence of high & low wind speed [J ] ． Smart Power ， 2024 ， 52 ( 7 )： 80 - 87 ．

黄肖琪，周羽生，周文晴，等．基于储能和无功优化的直驱机组海上风电场低电压穿越策略 [J ] ．电测与仪表， 2024 ， 61 ( 7 )： 57 - 64 ．

HUANG X Q ， ZHOU Y S ， ZHOU W Q ， et al ． Low voltage ride through strategy of the D-PMSG offshore wind power farm based on energy storage and reactive power optimization [J ] ． Electrical Measurement & Instrumentation ， 2024 ， 61 ( 7 )： 57 - 64 ．

高仕龙，陈武，杨勇，等．海上风电柔性低频输电换流器控制策略及试验 [J ] ．电工技术学报， 2024 ， 39 ( S1 )： 77 - 94 ．

GAO S L ， CHEN W ， YANG Y ， et al ． Control strategy and test of offshore wind flexible low frequency transmission converter [J ] ． Transactions of China Electrotechnical Society ， 2024 ， 39 ( S1 )： 77 - 94 ．

WANG Y ， ZOU R ， LIU F ， et al ． A review of wind speed and wind power forecasting with deep neural networks [J ] ． Applied Energy ， 2021 ， 304 ： 117766 ． doi: 10.1016/j.apenergy.2021.117766 http://dx.doi.org/10.1016/j.apenergy.2021.117766

CHEN J ， FU X ， ZHANG L ， et al ． A novel offshore wind power prediction model based on TCN-DANet-sparse transformer and considering spatio-temporal coupling in multiple wind farms [J ] ． Energy ， 2024 ， 308 ： 132899 ． doi: 10.1016/j.energy.2024.132899 http://dx.doi.org/10.1016/j.energy.2024.132899

HO C Y ， CHENG K S ， ANG C H ． Utilizing the random forest method for short-term wind speed forecasting in the coastal area of central Taiwan [J ] ． Energies ， 2023 ， 16 ( 3 )： 1374 - 1381 ． doi: 10.3390/en16031374 http://dx.doi.org/10.3390/en16031374

MA K K ， ZHANG W ， GUO Z ， et al ． A hybrid forecasting model for very short-term wind speed prediction based on secondary decomposition and deep learning algorithms [J ] ． Earth Science Informatics ， 2023 ， 16 ( 3 )： 2421 - 2438 ． doi: 10.1007/s12145-023-01044-1 http://dx.doi.org/10.1007/s12145-023-01044-1

KAZI F B A ， SADIA M ， LIU F ， et al ． Short-term electricity demand forecasting of Dhaka city using CNN with stacked BiLSTM [EB/OL ] ．( 2024-09-28 )[ 2024-10-15 ] ． https://doi.org/10.48550/arXiv.2406.06651 https://doi.org/10.48550/arXiv.2406.06651 ．

RAMADEVI B ， KASI V R ， BINGI K ． Hybrid LSTM-based fractional-order neural network for jeju island’s wind farm power forecasting [J ] ． Fractal and Fractional ， 2024 ， 8 ( 3 )： 149 - 157 ． doi: 10.3390/fractalfract8030149 http://dx.doi.org/10.3390/fractalfract8030149

YANG Z C ． Data-driven discrete cosine transform (DCT)-based modeling and simulation for hourly air humidity prediction [J ] ． Soft Computing ， 2024 ， 28 ( 1 )： 541 - 563 ． doi: 10.1007/s00500-023-08297-4 http://dx.doi.org/10.1007/s00500-023-08297-4

BAI S J ， KOLTER J Z ， VLADLEN K ． An empirical evaluation of generic convolutional and recurrent networks for sequence modeling speed forecasting [EB/OL ] ．( 2018-04-19 )[ 2024-10-18 ] ． https://doi.org/10.48550/arXiv.1803.01271 https://doi.org/10.48550/arXiv.1803.01271 ．

XIAO H ， HE X ， LI C ． Multi-step prediction of offshore wind power based on Transformer network and Huber loss [J ] ． International Journal of Electrical Power & Energy Systems ， 2024 ， 162 ： 110229 ． doi: 10.1016/j.ijepes.2024.110229 http://dx.doi.org/10.1016/j.ijepes.2024.110229

LI N ， DONG J ， LIU L ， et al ． A novel EMD and causal convolutional network integrated with Transformer for ultra short-term wind power forecasting [J ] ． International Journal of Electrical Power & Energy Systems ， 2023 ， 154 ： 109470 ． doi: 10.1016/j.ijepes.2023.109470 http://dx.doi.org/10.1016/j.ijepes.2023.109470

ZIM A H ， IQBAL A ， MALIK A ， et al ． TCNFormer： temporal convolutional network former for short-term wind speed forecasting [EB/OL ] ．( 2024-08-27 )[ 2024-1-18 ] ． https://doi.org/10.48550/arXiv.2408.15737 https://doi.org/10.48550/arXiv.2408.15737 ．

JIANG M ， ZENG P ， WANG K ， et al ． FECAM：frequency enhanced channel attention mechanism for time series forecasting [J ] ． Advanced Engineering Informatics ， 2023 ， 58 ： 102158 ． doi: 10.1016/j.aei.2023.102158 http://dx.doi.org/10.1016/j.aei.2023.102158

刘禹，陈戈华．基于正余弦分解和DCT域的干涉条纹去噪算法 [J ] ．中国激光， 2025 ， 52 ( 6 )： 0609001 ．

LIU Y ， CHEN G H ． Interference fringe denoising algorithm based on cosine and sine decomposition and DCT domain [J ] ． Chinese Laser ， 2025 ， 52 ( 6 )： 0609001 ．

YANG Z C ． Discrete cosine transform-based predictive model extended in the least-squares sense for hourly load forecasting [J ] ． IET Generation，Transmission & Distribution ， 2016 ， 10 ( 15 )： 3930 - 3939 ． doi: 10.1049/iet-gtd.2016.0689 http://dx.doi.org/10.1049/iet-gtd.2016.0689

刘伟，蔡东升，冯付勇，等．基于DCT-CNN-GRU的短期电力负荷预测研究 [J/OL ] ．电测与仪表， 1 - 11 [ 2024-10-30 ] ． https://link.cnki.net/urlid/23.1202.th.20240321.1323.002 https://link.cnki.net/urlid/23.1202.th.20240321.1323.002 ．

LIU W ， CAI D S ， FENG F Y ， et al ． Research on short-term power load forecasting based on DCT-CNN-GRU [J/OL ] ． Electric Measurement and Instrumentation ， 1 - 11 [ 2024-10-30 ] ． https://link.cnki.net/urlid/23.1202.th.20240321.1323.002 https://link.cnki.net/urlid/23.1202.th.20240321.1323.002 ．

ZHU J ， SU L ， LI Y ． Wind power forecasting based on new hybrid model with TCN residual modification [J ] ． Energy and AI ， 2022 ， 10 ： 100199 ． doi: 10.1016/j.egyai.2022.100199 http://dx.doi.org/10.1016/j.egyai.2022.100199

NGUYEN H K M ， PHAN Q-D ， WU Y K ， et al ． Multi-step wind power forecasting with stacked temporal convolutional network (S-TCN) [J ] ． Energies ， 2023 ， 16 ( 9 )： 3792 - 3799 ． doi: 10.3390/en16093792 http://dx.doi.org/10.3390/en16093792

KIM J ， OBREGON J ， PARK H ， et al ． Multi-step photovoltaic power forecasting using transformer and recurrent neural networks [J ] ． Renewable and Sustainable Energy Reviews ， 2024 ， 200 ： 114479 ． doi: 10.1016/j.rser.2024.114479 http://dx.doi.org/10.1016/j.rser.2024.114479

樊雅洁，王聪，张宏立，等．组合聚类和深度学习模型的风电场群风速预测 [J ] ．哈尔滨工业大学学报， 56 ( 12 )： 71 - 78 ．

FAN Y J ， WANG C ， ZHANG H L ， et al ． Wind speed prediction for wind farm clusters based on combined clustering and deep learning models [J ] ． Journal of Harbin Institute of Technology ， 56 ( 12 )： 71 - 78 ．

张昊立，张菁，倪建辉，等．引入注意力机制的LSTM-FCN海上风电功率预测 [J ] ．太阳能学报， 2024 ， 45 ( 6 )： 444 - 450 ．

ZHANG H L ， ZHANG J ， NI J H ， et al ． Offshore wind power prediction using LSTM-FCN with attention mechanism [J ] ． Journal of Solar Energy ， 2024 ， 45 ( 6 )： 444 - 450 ．

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Research on Large-Signal Stability Assessment Methods of DC Microgrids Based on Artificial Intelligence

Multi-User Short-Term Load Prediction Method Based on Frequency-Domain Enhanced Transformer

Review of Research on Wind-Solar-Hydro Complementary Power Generation Systems

Analysis of Key Technologies and Development Prospects for Renewable Energy-Powered Water Electrolysis for Hydrogen Production Based on Artificial Intelligence

Development Status and Suggestions of Hydrogen Production Technology by Offshore Wind Power

Related Author

LIU Sucheng

LI Long

HONG Tao

LIU Xiaodong

LUAN Li

SONG Liye

MENG Fanyu

CHEN Yulin

Related Institution

School of Electrical and Information Engineering, Anhui University of Technology, Maanshan

Key Lab of Power Electronics and Motion Control (Anhui University of Technology), Maanshan

Faculty of Electrical and Control Engineering, Liaoning Technical University

Hainan Institute of Zhejiang University

Chengdu Chengdian Electric Power Engineering Design Co., Ltd.

AI问答

⁰