电器与能效管理技术 ›› 2022, Vol. 0 ›› Issue (2): 1-5.doi: 10.16628/j.cnki.2095-8188.2022.02.001

• 研究与分析 •    下一篇

基于保护状态信息与凝聚层次聚类的电网故障定位算法

何森1, 陈义森2, 张厚荣1, 郑文坚1, 宋云海1, 尚佳宁1, 崔曼帝1, 王奇1, 常安1, 赖光霖1   

  1. 1.中国南方电网超高压输电公司 检修试验中心,广东 广州 510700
    2.华南理工大学,广东 广州 510640
  • 收稿日期:2021-09-28 出版日期:2022-02-28 发布日期:2022-03-31
  • 作者简介:何 森(1993—),男,主要从事输变电智能装备技术研究与应用工作。|陈义森(1996—),男,硕士研究生,研究方向为电力系统故障诊断与风险评估。|张厚荣(1989—),男,高级工程师,主要从事输变电在线监测技术研究。
  • 基金资助:
    国家自然科学基金(51677073);中央高校基本科研业务费重大产学研合作扶持专项(xzd1D2201280)

Power Grid Fault Location Algorithm Based on Protection State Information and Agglomerated Hierarchical Clustering

HE Sen1, CHEN Yisen2, ZHANG Hourong1, ZHENG Wenjian1, SONG Yunhai1, SHANG Jianing1, CUI Mandi1, WANG Qi1, CHANG An1, LAI Guanglin1   

  1. 1. Maintenance and Test Center of Extral High Voltage Company of China Southern Power Grid,Guangzhou 510700,China
    2. South China University of Technology,Guangzhou 510640,China
  • Received:2021-09-28 Online:2022-02-28 Published:2022-03-31

摘要:

针对现有电网故障定位算法对信息利用率不高而导致定位准确性及容错能力不足的问题,提出一种基于保护状态信息与凝聚层次聚类(AHC)的电网故障定位算法。首先,通过智能电子装置(IED)获取电网各断路器及保护设备的状态信息以构建故障特征向量;然后,应用AHC算法对特征向量进行聚类分析,依据元件与IED的关联方式求取故障元件;最后,与现有基于模糊C均值、改进谱聚类的电网故障定位算法进行对比。结果表明,所提算法具备不需要事先预判聚类簇数、故障定位结果更加稳定的优势,且在准确性以及对于IED信息的容错能力方面有所提高。

关键词: 电网故障定位, 凝聚层次聚类, 保护状态信息, 容错性

Abstract:

Aiming at the problem of insufficient accuracy and fault tolerance due to low information utilization of existing fault location algorithms,a fault location algorithm based on protection state information and Agglomerative Hierarchical Clustering (AHC) is proposed.Firstly,the state information of circuit breakers and protective devices in the power grid is obtained by Intelligent Electronic Device (IED) to construct the fault feature vector.Then,AHC algorithm is used to cluster the feature vectors,and the faulty elements are obtained according to the association between elements and IEDs.Finally,it is compared with the existing power grid fault location algorithms based on fuzzy C-means and improved spectral clustering.The results show that the proposed algorithm has the advantages of not necessary to predict the number of clusters in advance,more stable fault location results,and improved accuracy and fault tolerance for IED information.

Key words: power grid fault location, agglomeration hierarchical clustering, protection status information, fault tolerance

中图分类号: