物理数据融合的风电场功率预测

doi:10.11930/j.issn.1004-9649.202510007

摘要/Abstract

摘要：

现有风电功率预测大多依赖数据驱动或物理驱动的单一方法，少有研究将物理模型与数据驱动相结合，而这两种方法之间存在显著的互补潜力。建立了基于K均值聚类、经验模态分解与并行加权长短期记忆网络的数据驱动模型，并构建了融合物理驱动与数据驱动的风电场预测方法。以山西某风电场的实测数据为案例进行验证，所提物理数据融合方法的预测精度比纯数据驱动方法高21.67%，比基于经验尾流物理模型驱动方法高35.17%。该结果证实了物理数据融合方法在风电场功率预测中具有一定优越性及可靠性，能够满足风电场功率预测精度的要求。

关键词: 风电功率预测, 数据驱动, 物理数据融合

Abstract:

Power forecasting is a fundamental research topic in the wind power industry. Existing wind power forecasting methods predominantly rely on either data-driven or physics-driven approaches, with few studies combining physical models and data-driven techniques despite their significant complementary potential. A data-driven model is established using K-means clustering, empirical mode decomposition, and parallel weighted long short-term memory networks. A novel integrated approach combining physics-driven and data-driven methods was developed for wind farm forecasting. Validation using real-world data from a Chinese wind farm demonstrated that the proposed integrated method achieved 21.67% higher prediction accuracy than data-driven methods and 35.17% higher accuracy than physics-driven methods. These results confirm the superiority and reliability of physics-data fusion methods in wind farm ultra-short power forecasting.

Key words: wind power prediction, data-driven, physical-data fusion

赵军, 张世锋, 宋金鸽. 物理数据融合的风电场功率预测[J]. 中国电力, 2026, 59(5): 176-182.

ZHAO Jun, ZHANG Shifeng, SONG Jinge. Wind farm power forecasting by physical data fusion[J]. Electric Power, 2026, 59(5): 176-182.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.202510007

https://www.electricpower.com.cn/CN/Y2026/V59/I5/176

图/表 8

图 1 K-EMD-LSTM功率预测方法

Fig.1 K-EMD-LSTM power forecasting method

图 2 物理数据融合功率预测

Fig.2 Power prediction via physical data fusion

图 3 风电场机位

Fig.3 Wind farm turbine layout

图 4 风速、风向及温度数据

Fig.4 Wind speed, wind direction, and temperature data

图 5 不同数据驱动方法的功率预测误差柱状图

Fig.5 Bar chart of power forecasting errors for different data-driven methods

图 6 物理驱动、数据驱动及融合模型发电预测误差比较

Fig.6 Comparison of forecasting errors in physics-driven, data-driven, and fusion models for power generation

图 7 不同驱动模型预测误差指标

Fig.7 Bar chart of prediction error metrics for different driving models

表 1 不同方法在训练集上的误差比较

Table 1 Comparison of errors among different methods on the training set 单位：MW

模型	E_RMS	E_MA
物理驱动	18.024 8	12.648 9
数据驱动	24.197 6	14.302 5
融合模型	12.899 0	8.988 2

参考文献 35

1	王泽森, 王宣元, 孔帅皓, 等. 计及多时间尺度碳排放因子的虚拟电厂-配电网协同调度[J]. 中国电力, 2026, 59 (3): 14- 26.
	WANG Zesen, WANG Xuanyuan, KONG Shuaihao, et al. Coordinated dispatch of virtual power plant and distribution network considering multi-time scale carbon emission factors[J]. Electric Power, 2026, 59 (3): 14- 26.
2	贾东梨, 任昭颖, 刘科研, 等. 计及多种分布式能源的多端直流配电网故障电流计算方法[J]. 中国电力, 2026, 59 (1): 84- 96.
	JIA Dongli, REN Zhaoying, LIU Keyan, et al. Fault current calculation method for multi-terminal DC distribution networks considering multiple distributed generation[J]. Electric Power, 2026, 59 (1): 84- 96.
3	黄旭东, 胡彬, 赵志坚, 等. 构网型双馈风电机组机电耦合特性建模与轴系扭振分析[J]. 电力系统自动化, 2025, 49 (23): 68- 76.
	HUANG Xudong, HU Bin, ZHAO Zhijian, et al. Modeling of electromechanical coupling characteristics and analysis of shaft oscillation for grid-forming doubly-fed wind turbine[J]. Automation of Electric Power Systems, 2025, 49 (23): 68- 76.
4	郝露茜, 刘琳, 刘白杨, 等. 基于变分模态分解和麻雀搜索算法的双向长短期记忆网络的风电短期功率预测方法研究[J]. 湖南电力, 2024, 44 (3): 89- 95.
	HAO Luxi, LIU Lin, LIU Baiyang, et al. Research on short-term wind power prediction method of bidirectional long short-term memory based on variational mode decomposition and sparrow search algorithm[J]. Hunan Electric Power, 2024, 44 (3): 89- 95.
5	严新荣, 童跃平, 马奎超, 等. 风电机组叶片防除冰关键技术研究综述与展望[J]. 中国电力, 2026, 59 (3): 103- 113.
	YAN Xinrong, TONG Yueping, MA Kuichao, et al. Review and prospect of key technologies for anti/de-icing technologies of wind turbine blades[J]. Electric Power, 2026, 59 (3): 103- 113.
6	杨国华, 祁鑫, 贾睿, 等. 基于CEEMD-SE的CNN&LSTM-GRU短期风电功率预测[J]. 中国电力, 2024, 57 (2): 55- 61.
	YANG Guohua, QI Xin, JIA Rui, et al. Short-term wind power forecast based on CNN & LSTM-GRU model integrated with CEEMD-SE algorithm[J]. Electric Power, 2024, 57 (2): 55- 61.
7	朱东杰, 吕昆烨, 宋长虹, 等. 基于密度修正的风电功率曲线线性拟合模型[J]. 综合智慧能源, 2025, 47 (3): 84- 91.
	ZHU Dongjie, LYU Kunye, SONG Changhong, et al. Linear fitting model for wind power curves based on density correction[J]. Integrated Intelligent Energy, 2025, 47 (3): 84- 91.
8	张柏林, 李希德, 魏博, 等. 基于改进的场景分类和去粗粒化MCMC的风电出力模拟方法[J]. 电测与仪表, 2024, 61 (7): 41- 49, 122.
	ZHANG Bolin, LI Xide, WEI Bo, et al. Wind power output simulation method based on improved scene classification algorithm and coarse-grained MCMC[J]. Electrical Measurement & Instrumentation, 2024, 61 (7): 41- 49, 122.
9	程先龙, 张杰, 李思莹, 等. 基于VMD-BP-BiLSTM的短期风电功率预测[J]. 综合智慧能源, 2025, 47 (6): 20- 29.
	CHENG Xianlong, ZHANG Jie, LI Siying, et al. Short-term wind power prediction based on VMD-BP-BiLSTM[J]. Integrated Intelligent Energy, 2025, 47 (6): 20- 29.
10	司雅昕, 刘训强, 苏向敬, 等. 海上风电功率预测研究综述[J]. 浙江电力, 2025, 44 (11): 59- 71.
	SI Yaxin, LIU Xunqiang, SU Xiangjing, et al. A review of offshore wind power forecasting studies[J]. Zhejiang Electric Power, 2025, 44 (11): 59- 71.
11	黄学勤, 杨鹏举, 赵耀, 等. 考虑多维气象数据与时间影响的风电功率区间预测[J]. 浙江电力, 2026, 45 (1): 66- 77.
	HUANG Xueqin, YANG Pengju, ZHAO Yao, et al. Wind power interval forecasting based on multidimensional meteorological data and temporal effects[J]. Zhejiang Electric Power, 2026, 45 (1): 66- 77.
12	孟丹, 陈正洪, 许杨, 等. 极端天气对风电开发全过程的影响及应对策略[J]. 南方能源建设, 2025, 12 (2): 1- 14.
	MENG Dan, CHEN Zhenghong, XU Yang, et al. The impact of extreme weather on the entire process of wind power development and response strategies[J]. Southern Energy Construction, 2025, 12 (2): 1- 14.
13	彭穗, 李峰, 王彦峰, 等. 考虑极热极寒日的高比例风电电力系统备用容量规划[J]. 广东电力, 2025, 38 (10): 1- 13.
	PENG Sui, LI Feng, WANG Yanfeng, et al. Reserve capacity planning for high-proportion wind power systems considering extremely hot and cold days[J]. Guangdong Electric Power, 2025, 38 (10): 1- 13.
14	郑广宇, 辛征, 迟蔚然, 等. 基于偏最小二乘和改进时间卷积网络的风电场超短期发电功率预测[J]. 山东电力技术, 2025, 52 (12): 1- 16.
	ZHENG Guangyu, XIN Zheng, CHI Weiran, et al. Ultra-short term power generation prediction of wind farm based on partial least squares and improved time convolutional network[J]. Shandong Electric Power, 2025, 52 (12): 1- 16.
15	皇甫陈萌, 阮贺彬, 徐俊俊. 融合CEEMDAN-CNN-LSTM的风电机组多气象场景功率回归预测[J]. 综合智慧能源, 2025, 47 (9): 38- 50.
	HUANGFU Chenmeng, RUAN Hebin, XU Junjun. Power regression prediction for wind turbines in multi-meteorological scenarios based on CEEMDAN-CNN-LSTM integration[J]. Integrated Intelligent Energy, 2025, 47 (9): 38- 50.
16	苏向敬, 宇海波, 符杨, 等. 基于DALSTM和联合分位数损失的海上风电功率概率预测[J]. 中国电力, 2023, 56 (11): 10- 19.
	SU Xiangjing, YU Haibo, FU Yang, et al. Probabilistic forecasting of offshore wind power based on dual-stage attentional LSTM and joint quantile loss function[J]. Electric Power, 2023, 56 (11): 10- 19.
17	赵建红, 东星, 刘仲稳, 等. 基于模糊控制的混合储能平抑风电功率波动[J]. 湖南电力, 2023, 43 (6): 124- 131.
	ZHAO Jianhong, DONG Xing, LIU Zhongwen, et al. Hybrid energy storage based on fuzzy control stabilizes wind power fluctuations[J]. Hunan Electric Power, 2023, 43 (6): 124- 131.
18	GUO N Z, SHI K Z, LI B, et al. A physics-inspired neural network model for short-term wind power prediction considering wake effects[J]. Energy, 2022, 261, 125208.
19	VLADISLAVLEVA E, FRIEDRICH T, NEUMANN F, et al. Predicting the energy output of wind farms based on weather data: Important variables and their correlation[J]. Renewable Energy, 2013, 50, 236- 243.
20	FENG C, CUI M J, HODGE B M, et al. A data-driven multi-model methodology with deep feature selection for short-term wind forecasting[J]. Applied Energy, 2017, 190, 1245- 1257.
21	屈伯阳, 付立思. 基于NWP风速修正与VMD-DBO-DELM残差建模的风电功率预测研究[J]. 山东电力技术, 2025, 52 (2): 32- 45.
	QU Boyang, FU Lisi. Research on wind power prediction based on NWP wind speed correction and VMD-DBO-DELM residual modeling[J]. Shandong Electric Power, 2025, 52 (2): 32- 45.
22	BARAN S, LERCH S. Mixture EMOS model for calibrating ensemble forecasts of wind speed[J]. Environmetrics, 2016, 27 (2): 116- 130.
23	WANG C, ZHANG S H, LIAO P, et al. Wind speed forecasting based on hybrid model with model selection and wind energy conversion[J]. Renewable Energy, 2022, 196, 763- 781.
24	FAN H, ZHANG X M, MEI S W, et al. A Markov regime switching model for ultra-short-term wind power prediction based on toeplitz inverse covariance clustering[J]. Frontiers in Energy Research, 2021, 9, 638797.
25	RAISSI M, PERDIKARIS P, KARNIADAKIS G E. Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations[J]. Journal of Computational Physics, 2019, 378, 686- 707.
26	KARNIADAKIS G E, KEVREKIDIS I G, LU L, et al. Physics-informed machine learning[J]. Nature Reviews Physics, 2021, 3 (6): 422- 440.
27	袁飞, 夏德喜, 汪正军. 基于SCADA数据的风电机组群尾流效应计算与验证研究[J]. 智慧电力, 2023, 51 (7): 23- 30.
	YUAN Fei, XIA Dexi, WANG Zhengjun. Calculation and verification of wake effect on wind turbine based on SCADA data[J]. Smart Power, 2023, 51 (7): 23- 30.
28	YANG X Y, ZHANG Y F, YANG Y W, et al. Deterministic and probabilistic wind power forecasting based on bi-level convolutional neural network and particle swarm optimization[J]. Applied Sciences, 2019, 9 (9): 1794.
29	KENNEDY J, EBERHART R. Particle swarm optimization[C]//Proceedings of ICNN'95 - International Conference on Neural Networks. Perth, WA, Australia. IEEE, 2002: 1942-1948.
30	李振海, 李钰炎, 易志高, 等. 基于长短时记忆网络和轻梯度增强机的风电功率多步预测[J]. 湖南电力, 2023, 43 (6): 68- 75.
	LI Zhenhai, LI Yuyan, YI Zhigao, et al. Multi-step prediction of wind power based on long short-term memory network and light gradient boosting machine[J]. Hunan Electric Power, 2023, 43 (6): 68- 75.
31	ANGELIS D, SOFOS F, KARAKASIDIS T E. Artificial intelligence in physical sciences: symbolic regression trends and perspectives[J]. Archives of Computational Methods in Engineering, 2023, 30 (6): 3845- 3865.
32	张伟伟, 王旭, 寇家庆. 面向流体力学的多范式融合研究展望[J]. 力学进展, 2023, 53 (2): 433- 467.
	ZHANG Weiwei, WANG Xu, KOU Jiaqing. Prospects of multi-paradigm fusion methods for fluid mechanics research[J]. Advances in Mechanics, 2023, 53 (2): 433- 467.
33	WANG S, LI B, LI G Z, et al. Short-term wind power prediction based on multidimensional data cleaning and feature reconfiguration[J]. Applied Energy, 2021, 292, 116851.
34	CAMAL S, GIRARD R, FORTIN M, et al. A conditional and regularized approach for large-scale spatiotemporal wind power forecasting[J]. Sustainable Energy Technologies and Assessments, 2024, 65, 103743.
35	BARJASTEH A, GHAFOURI S H, HASHEMI M. A hybrid model based on discrete wavelet transform (DWT) and bidirectional recurrent neural networks for wind speed prediction[J]. Engineering Applications of Artificial Intelligence, 2024, 127, 107340.

[1]	王杰, 郑飞, 张鹏城, 陈露东, 高华, 蒙镜蓉. 基于数据驱动的高比例新能源配电网规划模型[J]. 中国电力, 2025, 58(3): 175-182.
[2]	计荣荣, 张淦锋, 王梦芝, 金颖杰, 李海, 黄若钒. 基于数据驱动的智能变电站信息高效校验技术[J]. 中国电力, 2025, 58(1): 100-106.
[3]	朱卫平, 汤奕, 魏兴慎, 刘增稷. 针对电力CPS数据驱动算法对抗攻击的防御方法[J]. 中国电力, 2024, 57(9): 32-43.
[4]	王阳, 马伟东, 刘洎溟, 王博石, 姚凯, 韩伟, 余娟. 基于关联规则与重构误差的二次系统故障检测方法[J]. 中国电力, 2024, 57(8): 159-167.
[5]	杨国华, 祁鑫, 贾睿, 刘一峰, 蒙飞, 马鑫, 邢潇文. 基于CEEMD-SE的CNN&LSTM-GRU短期风电功率预测[J]. 中国电力, 2024, 57(2): 55-61.
[6]	刘志伟, 马悦, 沙志成, 邵云姝, 牛远方, 董晓明, 王成福. 考虑新能源多重相关性的柔性配电网分布鲁棒优化策略[J]. 中国电力, 2024, 57(12): 97-108.
[7]	许竞, 赵铁军, 高小刚, 叶鞠, 孙玲玲. 高比例新能源电力系统调节资源灵活性不足风险分析[J]. 中国电力, 2024, 57(11): 129-138.
[8]	于宗超, 文明, 李湘华, 谢欣涛, 杨洪明. 融合群体智慧的分布式智能电网高效发展管理策略[J]. 中国电力, 2024, 57(10): 57-68.
[9]	姚力, 郑海峰, 单葆国, 谭显东, 许传龙, 徐志成. 基于数据驱动机会约束的发电企业电煤采购及库存优化模型[J]. 中国电力, 2023, 56(6): 176-184.
[10]	李鹏华, 宋卓然, 吴文传. 部分可观测条件下的配电网数据驱动最优潮流模型[J]. 中国电力, 2023, 56(12): 51-57.
[11]	黎冲, 王成辉, 王高, 鲁宗虎, 马成智. 基于数据驱动的锂离子电池健康状态估计技术[J]. 中国电力, 2022, 55(8): 73-86,95.
[12]	贾睿, 杨国华, 郑豪丰, 张鸿皓, 柳萱, 郁航. 基于自适应权重的CNN-LSTM&GRU组合风电功率预测方法[J]. 中国电力, 2022, 55(5): 47-56,110.
[13]	张健, 于浩, 梁建权, 王悦, 刘贺千. 基于数据驱动的多污染模式电能质量耦合性评估[J]. 中国电力, 2022, 55(11): 84-90.
[14]	李登峰, 杨旼才, 刘育明, 徐瑞林, 余霞, 李昭炯. 基于SVR数据驱动的发电机进相极限最优化求解方法[J]. 中国电力, 2021, 54(8): 136-143,153.
[15]	杨晓楠, 孙博, 郎燕生. 基于深度学习的特高压直流闭锁故障智能调度决策[J]. 中国电力, 2020, 53(6): 8-17.