Wind farm cluster power prediction based on graph deviation attention network with learnable graph structure and dynamic error correction during load peak and valley periods

被引:0
|
作者
Yang, Mao [1 ]
Guo, Yunfeng [1 ]
Huang, Tao [2 ]
Fan, Fulin [3 ,4 ]
Ma, Chenglian [1 ]
Fang, Guozhong [5 ]
机构
[1] Northeast Elect Power Univ, Key Lab Modern Power Syst Simulat & Control & Rene, Minist Educ, Jilin 132012, Peoples R China
[2] Politecn Torino, DOE, I-10129 Turin, Italy
[3] Harbin Inst Technol, Sch Elect Engn & Automat, Harbin 150001, Peoples R China
[4] Univ Strathclyde, Inst Energy & Environm, Glasgow G1 1XW, Scotland
[5] State Grid Corp China Northeast Branch, Changchun 130699, Peoples R China
关键词
Load peak and valley characteristics; Improved clustering distance; Learnable GDAN structure; Error dynamic correction; SVMD; Ultra-short-term power prediction; ELECTRICITY;
D O I
10.1016/j.energy.2024.133645
中图分类号
O414.1 [热力学];
学科分类号
摘要
The power prediction accuracy of wind farm cluster (WFC) seriously affects its consumption and the safe and stable operation of power system. The fluctuation of power between wind farms (WFs) significantly affects the wind power ultra-short-term prediction (WPUP) accuracy of WFC. In this regard, this paper proposes a graph deviation attention network (GDAN) considering improved clustering distance and learnable graph structure (LGS) for predicting and correcting the wind power of WFC. And used a weighted distance function combining sequence convergence smoothing effect and correlation to dynamically divide the WFC, and to learn and construct the graph structure. Proposed the GDAN with LGS to mine the convergence correlation of WF subclusters and establish power prediction model. Considering the characteristics of load peak and valley periods (LPVP), introduced a power correction coefficient to reduce the error, and used the successive variational mode decomposition (SVMD) to extract its key components to achieve power prediction and correction. The proposed method is applied to the WFC in Western Inner Mongolia, China. Compared with the comparison model before correction, the RMSE, MAE and MAPE are reduced by 4.27 %, 3.55 % and 17.92 % respectively, and the R2 and Pr are increased by 11.87 % and 9.88 % respectively.
引用
收藏
页数:26
相关论文
共 15 条
  • [1] Graph Neural Network-Based Wind Farm Cluster Speed Prediction
    Chen, Ruifeng
    Liu, Jiaming
    Wang, Fei
    Ren, Hui
    Zhen, Zhao
    2020 IEEE STUDENT CONFERENCE ON ELECTRIC MACHINES AND SYSTEMS (SCEMS 2020), 2020, : 982 - 987
  • [2] DHGAR: Multi-Variable-Driven Wind Power Prediction Model Based on Dynamic Heterogeneous Graph Attention Recurrent Network
    Xu, Mingrui
    Zhu, Ruohan
    Yu, Chengming
    Mi, Xiwei
    APPLIED SCIENCES-BASEL, 2025, 15 (04):
  • [3] Ultra-Short-Term Prediction of Wind Farm Cluster Power Based on Embedded Graph Structure Learning With Spatiotemporal Information Gain
    Yang, Mao
    Guo, Yunfeng
    Fan, Fulin
    IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, 2025, 16 (01) : 308 - 322
  • [4] The Short-Term Prediction of Wind Power Based on the Convolutional Graph Attention Deep Neural Network
    Xiao F.
    Ping X.
    Li Y.
    Xu Y.
    Kang Y.
    Liu D.
    Zhang N.
    Energy Engineering: Journal of the Association of Energy Engineering, 2024, 121 (02): : 359 - 376
  • [5] A short-term power prediction method based on numerical weather prediction correction and the fusion of adaptive spatiotemporal graph feature information for wind farm cluster
    Yang, Mao
    Han, Chao
    Zhang, Wei
    Fang, Guozhong
    Jia, Yunpeng
    EXPERT SYSTEMS WITH APPLICATIONS, 2025, 274
  • [6] Interpretable Power Output Prediction of Multiple Wind Turbines for Offshore Wind Farm Based on Multiple Spatio-temporal Attention Graph Neural Network Model
    Su X.
    Nie L.
    Li C.
    Mi Y.
    Fu Y.
    Dong Z.
    Dianli Xitong Zidonghua/Automation of Electric Power Systems, 2023, 47 (09): : 88 - 98
  • [7] Spatio-Temporal Graph Neural Network and Pattern Prediction Based Ultra-Short-Term Power Forecasting of Wind Farm Cluster
    Liu, Xiaoyan
    Zhang, Yiran
    Zhen, Zhao
    Xu, Fei
    Wang, Fei
    Mi, Zengqiang
    IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, 2024, 60 (01) : 1794 - 1803
  • [8] A centralized power prediction method for large-scale wind power clusters based on dynamic graph neural network
    Yang, Mao
    Wang, Da
    Zhang, Wei
    Yv, Xinnan
    ENERGY, 2024, 310
  • [9] Short-Term Power Forecasting of Wind Farm Cluster Based on Global Information Adaptive Perceptual Graph Convolution Network
    Yang, Mao
    Ju, Chaoyi
    Huang, Yutong
    Guo, Yunfeng
    Jia, Mengqi
    IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, 2024, 15 (03) : 2063 - 2076
  • [10] Dynamic spatio-temporal correlation and hierarchical directed graph structure based ultra-short-term wind farm cluster power forecasting method
    Wang, Fei
    Chen, Peng
    Zhen, Zhao
    Yin, Rui
    Cao, Chunmei
    Zhang, Yagang
    Duic, Neven
    APPLIED ENERGY, 2022, 323