1.东北电力大学 现代电力系统仿真控制与绿色电能新技术教育部重点实验室,吉林 吉林 132012
2.中国电力科学研究院有限公司 新能源与储能运行控制国家重点实验室,北京 100192
杨 茂(1982 —),男,教授,博士研究生导师,博士,研究方向为新能源功率预测(E-mail:yangmao820@163.com);
王 仲(1989 —),男,工程师,博士研究生,研究方向为新能源功率预测(E-mail:ele_wz@163.com);
王 达(1995 —),男,讲师,博士,通信作者,研究方向为新能源功率预测(E-mail:1610909212@qq.com)。
收稿:2025-11-26,
修回:2026-03-21,
网络首发:2026-05-20,
纸质出版:2026-10-10
移动端阅览
杨茂,王仲,王达等.基于时空特征复用的短期光伏功率预测方法[J].电力自动化设备,2026,46(10):124-132.
Yang Mao,Wang Zhong,Wang Da,et al.Short-term photovoltaic power prediction method based on spatiotemporal feature reuse[J].Electric Power Automation Equipment,2026,46(10):124-132.
杨茂,王仲,王达等.基于时空特征复用的短期光伏功率预测方法[J].电力自动化设备,2026,46(10):124-132. DOI: 10.16081/j.epae.202605006.
Yang Mao,Wang Zhong,Wang Da,et al.Short-term photovoltaic power prediction method based on spatiotemporal feature reuse[J].Electric Power Automation Equipment,2026,46(10):124-132. DOI: 10.16081/j.epae.202605006.
为提升光伏功率预测建模效率,构建多光伏电站短期功率同步预测框架。提出一种时空特征复用机制,在保留原有特征的前提下为预测模型决策扩充有效特征;构建一种动态时空图神经网络,在网络训练过程中自动调整节点连接方式,使图结构能够根据天气变化动态调整,增强对光伏出力时空相关性变化的感知能力,并通过改进的多任务学习机制优化网络输出层,挖掘各光伏电站预测任务之间的相关性,同时输出多座光伏电站的短期预测功率。基于我国某光伏集群数据进行实验分析,结果表明,日前光伏集群功率预测准确率达到91.42 %,比长短期记忆网络提升了1.37 %,验证了所提预测模型的有效性。
In order to improve the modeling efficiency of photovoltaic power prediction,a synchronous prediction framework of short-term power for multiple photovoltaic power plants is constructed. A spatiotemporal feature reuse mechanism is proposed,which expands the effective features for prediction model decision-making under the premise of retaining the original features. A dynamic spatiotemporal graph neural network is constructed,the node connection modes during the network training are automatically adjusted,which makes the graph structure can be dynamically adjusted according to weather change,and enhances the perception ability to the spatiotemporal correlation of photovoltaic output. The network output layer is optimized by an improved multi-task learning mechanism,which excavates the correlation between the prediction tasks of each photovoltaic power plant,and simultaneously outputs the short-term prediction power of multiple photovoltaic power plants. The experiment analysis based on the data of a photovoltaic cluster in China is carried out,and the results show that the accuracy of day-ahead power prediction for the photovoltaic cluster reaches 91.42 %,which is 1.37 % higher than that of the long short-term memory network,verifying the effectiveness of the proposed prediction model.
国家能源局 . 2025年可再生能源并网运行情况 [EB/OL]. [ 2025-10-20 ]. https:∥www.nea.gov.cn/20260212/742b8c6a-078347b0b39de676c05c5d58/c.html https://www.nea.gov.cn/20260212/742b8c6a-078347b0b39de676c05c5d58/c.html .
叶林 , 吕可欣 , 裴铭 , 等 . 考虑NWP传递熵因果验证的光伏功率预测误差溯源方法 [J]. 电力系统自动化 , 2025 , 49 ( 15 ): 110 - 122 .
Ye Lin , Lü Kexin , Pei Ming , et al . Error tracing method for photovoltaic power prediction considering causality verification of numerical weather prediction transfer entropy [J]. Automation of Electric Power Systems , 2025 , 49 ( 15 ): 110 - 122 .
文斌 , 章学勤 , 付文龙 , 等 . 基于二次模态分解重构及BiTCN-BiGRU模型的光伏短期发电功率预测 [J]. 电力系统保护与控制 , 2025 , 53 ( 18 ): 74 - 87 .
Wen Bin , Zhang Xueqin , Fu Wenlong , et al . Short-term PV power generation forecasting based on quadratic mode decomposition reconstruction and BiTCN-BiGRU model [J]. Power System Protection and Control , 2025 , 53 ( 18 ): 74 - 87 .
欧阳永健 , 缪希仁 , 林蔚青 , 等 . 基于分层有向图与动态时空相关性的小区域光伏超短期预测方法 [J]. 电网技术 , 2024 , 48 ( 6 ): 2458 - 2468 .
Ouyang Yongjian , Miao Xiren , Lin Weiqing , et al . Small area PV ultra-short-term prediction method using stratified digraph and dynamic spatial-temporal correlation [J]. Power System Technology , 2024 , 48 ( 6 ): 2458 - 2468 .
李嵩 , 夏成军 , 刘译夫 , 等 . 基于KAN和误差时间序列自相关性特征的多时间尺度光伏功率预测 [J]. 电力系统保护与控制 , 2025 , 53 ( 24 ): 176 - 187 .
Li Song , Xia Chengjun , Liu Yifu , et al . Multi-time scale PV power forecasting based on KAN and error time series autocorrelation features [J]. Power System Protection and Control , 2025 , 53 ( 24 ): 176 - 187 .
许文举 , 张孝顺 , 郭正勋 , 等 . 基于图变换神经网络的电网用户侧节点碳排放因子预测算法 [J]. 电网技术 , 2024 , 48 ( 12 ): 4980 - 4988 .
Xu Wenju , Zhang Xiaoshun , Guo Zhengxun , et al . Prediction algorithm of carbon emission factor of power grid user-side nodes based on variant graph neural network [J]. Power System Technology , 2024 , 48 ( 12 ): 4980 - 4988 .
Wu T , Hu R F , Zhu H Y , et al . Combined IXGBoost-KELM short-term photovoltaic power prediction model based on multi-dimensional similar day clustering and dual decomposition [J]. Energy , 2024 , 288 : 129770 .
Tan L , Kang R X , Xia J M , et al . Application of multi-source data fusion on intelligent prediction of photovoltaic power [J]. Solar Energy , 2024 , 277 : 112706 .
Cui S H , Lyu S P , Ma Y Z , et al . Improved Informer PV po-wer short-term prediction model based on weather typing and AHA-VMD-MPE [J]. Energy , 2024 , 307 : 132766 .
段宏 , 郭成 , 孙海东 , 等 . 基于DDTW聚类和SK TCN-GC BiGRU的分布式光伏短期功率预测 [J]. 智慧电力 , 2025 , 53 ( 4 ): 71 - 80 .
Duan Hong , Guo Cheng , Sun Haidong , et al . Short-term power prediction for distributed photovoltaic systems based on DDTW clustering and SK TCN-GC BiGRU network [J]. Smart Power , 2025 , 53 ( 4 ): 71 - 80 .
Yang X Y , Yang L , Li Y K , et al . A new method of photo-voltaic clusters power prediction based on Informer considering time-frequency analysis and convergence effect [J ]. Electric Power Systems Research , 2025 , 238 : 111049 .
Keddouda A , Ihaddadene R , Boukhari A , et al . Solar photovol-taic power prediction using artificial neural network and multi-ple regression considering ambient and operating conditions [J]. Energy Conversion and Management , 2023 , 288 : 117186 .
Fu L , Yang Y L , Yao X L , et al . A regional photovoltaic out-put prediction method based on hierarchical clustering and the mRMR criterion [J]. Energies , 2019 , 12 ( 20 ): 3817 .
Qiu L H , Ma W T , Feng X Y , et al . A hybrid PV cluster power prediction model using BLS with GMCC and error cor-rection via RVM considering an improved statistical upsca-ling technique [J]. Applied Energy , 2024 , 359 : 122719 .
Ouyang J , Chu L D , Chen X L , et al . A K -means cluster divi-sion of regional photovoltaic power stations considering the con-sistency of photovoltaic output [J]. Sustainable Energy,Grids and Networks , 2024 , 40 : 101573 .
Tahir M F , Yousaf M Z , Tzes A , et al . Enhanced solar photo-voltaic power prediction using diverse machine learning algo-rithms with hyperparameter optimization [J]. Renewable and Sustainable Energy Reviews , 2024 , 200 : 114581 .
沈仲恺 , 宗星辰 , 万灿 , 等 . 基于光伏功率概率预测的新能源配电系统节点电压不确定性量化 [J]. 电力自动化设备 , 2026 , 46 ( 1 ): 31 - 39 .
Shen Zhongkai , Zong Xingchen , Wan Can , et al . Photovoltaic power probability forecasting based uncertainty quantification of node voltage in distribution system with renewable energy [J]. Electric Power Automation Equipment , 2026 , 46 ( 1 ): 31 - 39 .
Xiao Z N , Huang X Q , Liu J , et al . A novel method based on time series ensemble model for hourly photovoltaic power prediction [J]. Energy , 2023 , 276 : 127542 .
Hu Z H , Gao Y , Ji S Y , et al . Improved multistep ahead photovoltaic power prediction model based on LSTM and self-attention with weather forecast data [J]. Applied Energy , 2024 , 359 : 122709 .
Li N Q , Li L H , Zhang F , et al . Research on short-term photo-voltaic power prediction based on multi-scale similar days and ESN-KELM dual core prediction model [J]. Energy , 2023 , 277 : 127557 .
Li Y K , Huang W , Lou K Y , et al . Short-term PV power prediction based on meteorological similarity days and SSA-BiLSTM [J]. Systems and Soft Computing , 2024 , 6 : 200084 .
邹小龙 , 温步瀛 , 王怀远 . 融合物理信息的光伏发电功率短期预测方法 [J]. 智慧电力 , 2025 , 53 ( 12 ): 59 - 67 .
Zou Xiaolong , Wen Buying , Wang Huaiyuan . A physics-informed short-term forecasting method for photovoltaic power generation [J]. Smart Power , 2025 , 53 ( 12 ): 59 - 67 .
Zhang R Y , Ma H , Hua W , et al . Data-driven photovoltaic gene-ration forecasting based on a Bayesian network with spatial-temporal correlation analysis [J]. IEEE Transactions on Indus-trial Informatics , 2020 , 16 ( 3 ): 1635 - 1644 .
Miraftabzadeh S M , Colombo C G , Longo M , et al . A day-ahead photovoltaic power prediction via transfer learning and deep neural networks [J]. Forecasting , 2023 , 5 ( 1 ): 213 - 228 .
Li F Y , Zheng H F , Li X M . A novel hybrid model for multi-step ahead photovoltaic power prediction based on conditio-nal time series generative adversarial networks [J]. Renewable Energy , 2022 , 199 : 560 - 586 .
赵如意 , 王晓辉 , 郑碧煌 , 等 . 基于特征优化和混合改进灰狼算法优化BiLSTM网络的短期光伏功率预测 [J]. 电网技术 , 2025 , 49 ( 1 ): 209 - 222 .
Zhao Ruyi , Wang Xiaohui , Zheng Bihuang , et al . Short-term photovoltaic power prediction based on feature optimization and hybrid improved grey wolf algorithm-optimized BiLSTM network [J]. Power System Technology , 2025 , 49 ( 1 ): 209 - 222 .
颜俊 , 贺伟 , 郭创新 , 等 . 基于生成对抗网络与Transformer的多尺度光伏出力预测 [J]. 电力自动化设备 , 2025 , 45 ( 11 ): 25 - 35 .
Yan Jun , He Wei , Guo Chuangxin , et al . Multi-scale photovoltaic power output prediction based on generative adversarial network and Transformer [J]. Electric Power Automation Equip-ment , 2025 , 45 ( 11 ): 25 - 35 .
Simeunovic J , Schubnel B , Alet P J , et al . Spatio-temporal graph neural networks for multi-site PV power forecasting [J]. IEEE Transactions on Sustainable Energy , 2022 , 13 ( 2 ): 1210 - 1220 .
Wang K , Wang J Z , Li Z W , et al . An integrated sparse gated graph density network based on transfer learning for multi-site probabilistic forecasting of renewable energy [J]. IEEE Transactions on Sustainable Energy , 2025 , 16 ( 1 ): 673 - 685 .
陈昕钰 , 余光正 , 陈汝斯 , 等 . 考虑沙戈荒地区复杂气象因素的超短期光伏功率预测 [J]. 电力系统自动化 , 2026 , 50 ( 2 ): 83 - 92 .
Chen Xinyu , Yu Guangzheng , Chen Rusi , et al . Ultra-short-term photovoltaic power prediction considering complex meteorological factors in Gobi Desert and other arid areas [J]. Automation of Electric Power Systems , 2026 , 50 ( 2 ): 83 - 92 .
Zhang P , Zhang B , Yin J L , et al . County-level distributed PV day-ahead power prediction based on grey correlation analysis and transformer-GCAN model [J]. IEEE Transactions on Sus-tainable Energy , 2026 , 17 ( 1 ): 709 - 717 .
杨茂 , 孟庆虎 , 王勃 . 基于电量-功率的多模态映射的长预见期光伏集群功率预测 [J]. 电力自动化设备 , 2024 , 44 ( 11 ): 60 - 65,87 .
Yang Mao , Meng Qinghu , Wang Bo . Power forecast of photovoltaic cluster in long forecast period based on electricity-power multimodal mapping [J]. Electric Power Automation Equipment , 2024 , 44 ( 11 ): 60 - 65,87 .
Zhang M , Zhen Z , Liu N , et al . Optimal graph structure based short-term solar PV power forecasting method considering sur-rounding spatio-temporal correlations [J]. IEEE Transactions on Industry Applications , 2023 , 59 ( 1 ): 345 - 357 .
Zhang S X , Liu Z T , An T , et al . Hot rolled prognostic app-roach based on hybrid Bayesian progressive layered extraction multi-task learning [J]. Expert Systems with Applications , 2024 , 249 : 123763 .
0
浏览量
0
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621