电器与能效管理技术 ›› 2026, Vol. 0 ›› Issue (3): 23-31.doi: 10.16628/j.cnki.2095-8188.2026.03.004

• 研究与分析 • 上一篇    下一篇

基于数据关联图谱的电网调控数据自动修复方法研究

谢琳, 陶蕾, 王佳琪, 李大鹏, 叶瑞丽, 张周杰, 李嘉睿   

  1. 中国电力科学研究院有限公司 电力调度自动化技术研究与系统评价北京市重点实验室, 北京 100192
  • 收稿日期:2025-11-27 出版日期:2026-03-30 发布日期:2026-04-20
  • 作者简介:谢 琳(1992—),女,工程师,硕士,主要从事电网调度自动化、调控大数据及人工智能的应用研究。|陶 蕾(1989—),女,高级工程师,硕士,主要从事电网调度自动化、调控大数据及人工智能的应用研究。|王佳琪(1990—),女,工程师,硕士,主要从事电网调度自动化研究工作。
  • 基金资助:
    中国电力科学研究院有限公司自筹项目(52420023008D)

Research on Power Grid Dispatching and Control Data Automatic Repair Methods Based on Data Association Graphs

XIE Lin, TAO Lei, WANG Jiaqi, LI Dapeng, YE Ruili, ZHANG Zhoujie, LI Jiarui   

  1. China Electric Power Research Institute, Beijing Key Laboratory of Research and System Evaluation of Power Dispatching Automation Technology, Beijing 100192, China
  • Received:2025-11-27 Online:2026-03-30 Published:2026-04-20

摘要:

随着新型电力系统建设的深入推进,高质量数据已成为提升大电网协同控制与调度精益化管理水平的关键支撑。针对电网跨层级协同调度中海量多源异构数据异常修复难题,提出一种基于数据关联图谱的智能修复方法。通过融合调度对象关系、时空关联与数据血缘,构建全景电网调控数据关联图谱,实现数据细粒度全生命周期刻画。在此基础上,提出异常数据双向追踪算法,实现异常高效识别与准确定位;进而构建融合规则、模型与知识推理的多策略协同修复方法,结合修复影响度评估生成最优修复方案。实验结果表明,该方法在修复效率与准确性方面均具显著优势,可有效提升电网调控数据质量,为智能电网安全稳定运行提供可靠数据支撑。

关键词: 电网调控数据, 数据血缘, 全景数据关联图谱, 多策略协同, 数据修复

Abstract:

With the advancement of the new power system,high-quality data has become a crucial support for enhancing the coordinated control of large power grids and the lean management of dispatching.Aiming at the challenge of repairing abnormal massive,multi-source and heterogeneous data in cross-level collaborative power grid dispatching,an intelligent data repair method based on a data association graph is proposed.By integrating dispatching object relationships,spatio-temporal correlations,and data lineage,a panoramic power grid regulation and control data association graph is constructed,achieving the,fine-grained depiction of data throughout its entire life cycle.On this basis,a bidirectional tracking algorithm for abnormal data is proposed to enable efficient identification and accurate localization of anomalies.Furthermore,a multi-strategy collaborative data repair method integrating rules,models,and knowledge reasoning is developed,which combines repair impact assessment algorithm to generate the optimal repair strategy.Experimental results demonstrate that this method offers significant advantages in both repair efficiency and accuracy,effectively improving the quality of power grid regulation and control data and providing reliable data support for the safe and stable operation of smart grids.

Key words: power grid dispatching and control data, data lineage, panoramic data association graphs, multi-strategy collaboration, data repair

中图分类号: