基于特征分解与Bi-LSTM-Attention模型的风向预测

马良玉; 段晓冲; 胡景琛; 黄日灏; 程泽龙; 段新会

基于特征分解与Bi-LSTM-Attention模型的风向预测

1. 华北电力大学自动化系;
2. 保定市综合能源系统状态检测与优化调控重点实验室;
3. 保定华仿科技股份有限公司

基金项目:

河北省中央引导地方科技发展资金项目(226Z2103G)

详细信息

作者简介:
马良玉(1972—),男,教授,研究方向为智能技术在电站建模、优化控制与故障诊断中的应用;段新会(1969—)男,教授级高级工程师,研究方向为发电领域系统设备及控制系统的建模、仿真与控制

通讯作者:
段晓冲(2000—),男,硕士研究生,研究方向为发电过程建模、仿真与控制

中图分类号: TM614;TP183
计量
- 文章访问数: 0
- HTML全文浏览量: 0
- PDF下载量: 0
出版历程
- 收稿日期: 2024-04-11
- 刊出日期: 2024-08-27

Wind Direction Prediction Based on Feature Decomposition and Bi-LSTM-Attention Model

1. Department of Automation, North China Electric Power University;
2. Baoding Key Laboratory of State Detection and Optimization for Integrated Energy System;
3. Baoding SinoSimu Technology Co., Ltd.

摘要

摘要: 为便于精准控制风电机组的偏航角度、充分利用风能提高机组发电量，提出一种基于历史数据深度学习的风向超短期预测方法。首先利用变分模态分解将风向数据分解成多个子序列，考虑分解后的残差分量仍保留大量信号特征，进一步采用自适应噪声完备集合经验模态分解方法对残差分量进行二次分解。在此基础上，结合风速、环境温度等特征，利用具有注意力机制的双向长短期记忆网络对风向进行超短期预测。采用河北某风电场SCADA真实数据，对风向进行5 min的超短期预测实验，并与其他方法进行对比，结果表明所提方法具有更好的风向预测效果。
- 风向预测 /
- 变分模态分解 /
- CEEMDAN /
- 双向长短期记忆网络 /
- 注意力机制
Abstract: In order to accurately control the yaw angle of the wind turbine and fully utilize wind energy to increase the power generation of the unit, a deep learning based wind direction ultra-short-term prediction method based on historical data is proposed. Firstly, variational modal decomposition（VMD）is used to decompose wind direction data into multiple subsequences. Considering that the residual component after decomposition still retains a large number of signal features, adaptive noise complete set empirical mode decomposition（CEEMDAN） is further used to perform secondary decomposition for the residual component. On this basis, combined with features such as wind speed and environmental temperature, a bidirectional long short-term memory network with attention mechanism（Bi-LSTM-Attention） is used to predict wind direction in ultra-short-term. Using real SCADA data from a wind farm in Hebei, 5-minute wind direction prediction experiments are conducted, and compared with other methods, the results show that better prediction performance can be got with the proposed method.
- wind direction prediction /
- VMD /
- CEEMDAN /
- Bi-LSTM /
- attention mechanism

HTML全文

参考文献(14)

[1]	任曦骏,朱刘柱,谢道清,等.可再生能源参与电力现货市场关键问题的研究[J].现代电力, 2022, 39(2):203-211.REN XIJUN, ZHU LIUZHU, XIE DAOQING, et al.Discussion on key issues of renewable energy participation in the electricity spot market and its application on a case study[J]. Modern Electric Power,2022, 39(2):203-211(in Chinese).
[2]	王奕洲.一种基于风况预测的优化偏航策略[D].北京:华北电力大学, 2022.WANG YIZHOU. An optimized yaw strategy based on wind condition prediction[D]. Beijing:North China Electric Power University, 2022(in Chinese).
[3]	苏智超.基于深度学习的风电机组风向预测方法研究[J].电力与能源, 2023, 44(3):239-246.SU ZHICHAO. Study on wind direction prediction method for wind turbine based on deep learning[J].Power&Energy, 2023, 44(3):239-246(in Chinese).
[4]	王芳,李泽兰,谢卫才,等.基于卡尔曼滤波和时间序列相结合的变桨偏航控制方法[J].科技与创新,2023(15):1-4.
[5]	唐振浩,赵赓楠,曹生现,等.一种基于数据解析的混合风向预测算法[J].太阳能学报, 2021, 42(9):349-356.TANG ZHENHAO, ZHAO GENGNAN, CAO SHENGXIAN, et al. A data analystic based hybrid wind direction prediction algorithm[J]. Acta Energiae Solaris Sinica, 2021, 42(9):349-356(in Chinese).
[6]	辛欣瑶,汪建文,赵爽.基于统计方法的自然风风向预测模型[J].工程热物理学报, 2021, 42(6):1438-1445.XIN XINYAO, WANG JIANWEN, ZHAO SHUANG.Prediction model of natural wind direction based on statistical method[J]. Journal of Engineering Thermophysics, 2021, 42(6):1438-1445(in Chinese).
[7]	PATTANAIK B, MANIKANDAN S, PAULDOSS S P,et al. Prediction of the speed and wind direction using machine learning[C]//Journal of Physics:Conference Series. IOP Publishing, 2021, 1964(4):042064.
[8]	张琰妮,史加荣,李津,等.融合残差与VMD-ELMLSTM的短期风速预测[J].太阳能学报, 2023, 44(9):340-347.ZHANG YANNI, SHI JIARONG, LI JIN, et al.Short-term wind speed prediction based on residual and VMD-ELM-LSTM[J]. Acta Energiae Solaris Sinica,2023, 44(9):340-347(in Chinese).
[9]	韩星晖,何永玲,蒙占彬,等.基于CNN-LSTMAttention的短期风向预测研究[J].广西科学院学报,2023, 39(2):192-198.HAN XINGHUI, HE YONGLING, MENG ZHANBIN,et al. Short-term wind direction prediction based on CNN-LSTM Attention[J]. Journal of Guangxi Academy of Sciences, 2023, 39(2):192-198(in Chinese).
[10]	LIU B T, XIE Y Q, WANG K, et al. Short-term multi-step wind direction prediction based on OVMD quadratic decomposition and LSTM[J]. Sustainability,2023, 15(15):11746.
[11]	DRAGOMIRETSKIY K, ZOSSO D. Variational mode decomposition[J]. IEEE Transactions on Signal Processing, 2013, 62(3):531-544.
[12]	TORRES M E, COLOMINAS M A, SCHLOTTHAUER G, et al. A complete ensemble empirical mode decomposition with adaptive noise[C]//2011 IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP). Prague, Czech Republic:IEEE, 2011:4144-4147.
[13]	AGGARWAL C C. Neural Networks and Deep Learning[M]. Cham:Springer, 2018.
[14]	马良玉,段晓冲,胡景琛,等.基于LOF-DBSCAN与优化XGBoost的风电机组异常状态预警[J/OL].华北电力大学学报(自然科学版):1-9[2023-10-27]. http://kns.cnki.net/kcms/detail/13.1212.TM.20230829.1413.004.html.MA" target="_blank"> http://kns.cnki.net/kcms/detail/13.1212.TM.20230829.1413.004.html.MA LIANGYU, DUAN XIAOCHONG, HU JINGCHEN, et al. Anomaly warning of wind turbine based on LOF-DBSCAN and Optuna optimized XGBoost[J/OL]. Journal of North China Electric Power University(Natural Science Edition):1-9[2023-10-27]. http://kns.cnki.net/kcms/detail/13.1212.TM.20230829.1413.004.html(in Chinese).