电工技术学报  2022, Vol. 37 Issue (23): 6145-6156    DOI: 10.19595/j.cnki.1000-6753.tces.211486
电力系统与综合能源 |
用于未来态预测的电网运行断面时空相似性挖掘
顾雪平1, 刘彤1, 李少岩1, 王铁强2, 杨晓东2
1.华北电力大学电气与电子工程学院 保定 071003;
2.国网河北省电力公司 石家庄 050021
Temporal and Spatial Similarity Mining of Power Grid Running Section for Future State Prediction
Gu Xueping1, Liu Tong1, Li Shaoyan1, Wang Tieqiang2, Yang Xiaodong2
1. School of Electrical & Electronic Engineering North China Electric Power University Baoding 071003 China;
2. State Grid Hebei Electric Power Company Shijiazhuang 050021 China
全文: PDF (2031 KB)   HTML
输出: BibTeX | EndNote (RIS)      
摘要 运行状态有效预测可为电网风险预判与优化调控提供数据基础,对保障系统安全运行具有重要意义。该文提出一种用于未来态预测的电网运行断面时空相似性挖掘方法。首先,采用图表示学习算法对电网拓扑及其属性信息进行深层次无监督学习,提取表征运行断面空间特征的属性向量;然后,利用滑动时间窗算法将历史运行断面对应的空间特征向量按照不同时段划分到多个窗口;最后,从微观和宏观两角度计算不同窗口间对应样本相似性,获取与当前时段内断面最相似的一组连续断面,并将该组历史断面的后续时刻断面作为当前电网运行未来状态的参考。通过新英格兰10机39节点系统和实际电网算例进行验证,结果表明所提方法能够有效提取最相似的历史断面,进而实现对未来状态的辅助预测。
服务
把本文推荐给朋友
加入我的书架
加入引用管理器
E-mail Alert
RSS
作者相关文章
顾雪平
刘彤
李少岩
王铁强
杨晓东
关键词 运行状态预测图表示学习算法运行断面空间特征时段划分相似性匹配    
Abstract:Effective prediction of operation state can provide a data basis for power grid risk prediction and optimal regulation, which is of great significance for ensuring the system safe operation. This paper proposes a method for mining the temporal and spatial similarity of power grid running section for future state prediction. Firstly, the graph representation learning algorithm was used to deeply unsupervised learn the power grid topology and its attribute information, and the attribute vector representing the spatial features of running section was extracted. Further, the spatial feature vectors corresponding to the historical running sections was divided into multiple windows according to different time periods through the sliding time window algorithm. Finally, the similarity of corresponding samples between different windows was calculated from the microscopic and macroscopic perspectives to obtain a group of continuous sections that were most similar to the sections in the current period. The subsequent section of the most similar historical section was used as a reference for the future state of the current power grid. The results validated by IEEE 39-bus system and practical power grid simulation examples show that the proposed method can effectively match the most similar historical sections, then realize the auxiliary prediction of the future operation state.
Key wordsOperation state prediction    graph representation learning algorithm    running section spatial feature    period division    similarity matching   
收稿日期: 2021-09-23     
PACS: TM732  
基金资助:国家电网公司科技项目资助(SGTYHT/17-JS-199)
通讯作者: 刘 彤 女,1996年生,博士研究生,研究方向为人工智能技术及其在电力系统中的应用、电力系统安全稳定评估与控制。E-mail:tongliu_1996@163.com   
作者简介: 顾雪平 男,1964年生,教授,博士生导师,研究方向为电力系统安全稳定评估与控制、电力系统安全防御与恢复控制、人工智能技术及其在电力系统中的应用等。E-mail:xpgu@ncepu.edu.cn
引用本文:   
顾雪平, 刘彤, 李少岩, 王铁强, 杨晓东. 用于未来态预测的电网运行断面时空相似性挖掘[J]. 电工技术学报, 2022, 37(23): 6145-6156. Gu Xueping, Liu Tong, Li Shaoyan, Wang Tieqiang, Yang Xiaodong. Temporal and Spatial Similarity Mining of Power Grid Running Section for Future State Prediction. Transactions of China Electrotechnical Society, 2022, 37(23): 6145-6156.
链接本文:  
https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.211486          https://dgjsxb.ces-transaction.com/CN/Y2022/V37/I23/6145