中国电力 ›› 2024, Vol. 57 ›› Issue (12): 2-16.DOI: 10.11930/j.issn.1004-9649.202410093

• 面向新型电力系统的源荷预测技术 • 上一篇    下一篇

从感知-预测-优化综述图神经网络在电力系统中的应用

李卓1(), 王胤喆1(), 叶林1(), 罗雅迪2, 宋旭日2, 张振宇3   

  1. 1. 中国农业大学 信息与电气工程学院,北京 100083
    2. 中国电力科学研究院有限公司,北京 100192
    3. 国家电力调度控制中心,北京 100031
  • 收稿日期:2024-10-29 出版日期:2024-12-28 发布日期:2024-12-27
  • 作者简介:李卓(1994—),女,博士研究生,从事电力系统自动化、新能源发电技术研究,E-mail:lizhuo94@cau.edu.cn
    王胤喆(2000—),男,硕士研究生,从事电力系统自动化、新能源发电技术研究,E-mail:wangyinzhe@cau.edu.cn
    叶林(1968—),男,通信作者,博士,教授,博士生导师,从事电力系统自动化、新能源发电技术研究,E-mail:yelin@cau.edu.cn
  • 基金资助:
    国家电网有限公司科技项目(5108-202218280A-2-294-XG);国家自然科学基金新型电力系统联合基金资助项目(U22B20117)。

The Application of Graph Neural Networks in Power Systems from Perspective of Perception-Prediction-Optimization

Zhuo LI1(), Yinzhe WANG1(), Lin YE1(), Yadi LUO2, Xuri SONG2, Zhenyu ZHANG3   

  1. 1. College of Information and Electrical Engineering, China Agricultural University, Beijing 100083, China
    2. China Electric Power Research Institute, Beijing 100192, China
    3. State Grid National Power Dispatching and Control Center, Beijing 100031, China
  • Received:2024-10-29 Online:2024-12-28 Published:2024-12-27
  • Supported by:
    This work is supported by Science and Technology Project of SGCC (No.5108-202218280A-2-294-XG), Joint Fund for New Power System of National Natural Science Foundation of China (No.U22B20117).

摘要:

随着新型电力系统发电侧、输电侧和用电侧不确定性的日益增加,电力系统拓扑结构关系逐渐复杂、规模程度不断升级。常规欧式空间数据解析方法在表征多源异构和非规则的拓扑结构关系时,往往呈现性能较差、准确度不高的问题。图神经网络(graph neural networks,GNNs)能够捕捉到不同节点和边之间的复杂依赖关系,并有效挖掘非欧式空间数据结构中的时空特征,适用于复杂电力系统拓扑结构关系的感知与建模。针对于此,基于前人的研究进展,介绍了GNNs的定义和特点,并分析了GNNs不同变体的特点及其优势。然后,归纳和总结了GNNs在电力系统状态感知、预测、图潮流计算等方面的应用现状,从感知-预测-优化角度探讨了GNNs与新型电力系统的适配关系。最后,针对GNNs潜在的问题难点和未来可行的发展方向进行了总结和展望。

关键词: 新型电力系统, 不确定性, 图神经网络, 状态感知, 预测, 图潮流计算

Abstract:

With the increasing uncertainty of the generation, transmission, and consumption sides in new power systems, the complexity and scale of power system topology relationship are continuously growing. Conventional data analysis methods for Euclidean space often exhibit poor performance and low accuracy when representing the topological structures relationship with multi-source heterogeneous and irregular characteristics. Graph Neural Networks (GNNs) are capable of capturing complex dependency relationship between different nodes and edges, and effectively mining spatiotemporal features in non-Euclidean data structures, are therefore suitable for the perception and modeling of complex power system topologies. In this context, this paper builds upon previous research progress, providing the definition and characteristics of GNNs, and discussing the unique features and advantages of different variants GNNs. After that, it summarizes the current applications of GNNs in power system state perception, prediction, and graph-based power flow calculation, aiming to explore the suitability of GNNs for new power systems from the perception-prediction-optimization perspectives. Finally, a summary and outlook on the potential challenges and future development directions for GNNs are provided.

Key words: new power systems, uncertainty, graph neural networks, state perception, prediction, graph-based power flow calculation