基于多暂态特征量聚类的配电网接地故障区段定位方法  

A method of grounding fault location in distribution network based on multi-transient characteristic cluster

作　　者：林骏捷 林佳壕 郭谋发[1] LIN Junjie;LIN Jiahao;GUO Moufa(College of Electrical Engineering and Automation,Fuzhou University,Fuzhou 350108)

机构地区：[1]福州大学电气工程与自动化学院,福州350108

出　　处：《电气技术》2023年第5期16-22,共7页Electrical Engineering

摘　　要：谐振接地系统发生单相接地故障时,线路健全区段与故障区段的零序电压电流的伏安特性和零序功率方向存在差异。单相接地故障的影响因素多,且呈现随机性,采用单一接地故障特征量的区段定位方法准确率不高。为此,本文提出一种基于多暂态特征量聚类的接地故障区段免阈值定位方法。首先求取发生故障后线路各区段首半波暂态零序电压导数与暂态零序电流之间的伏安特性特征向量和零序功率特征向量,再对两特征量做归一化处理并融合得到综合特征向量,最后由免阈值设定的模糊C均值聚类算法判别健全区段与故障区段。通过仿真案例验证了所提方法的有效性。When a single-line-to-ground fault(SLGF)occurs in power distribution system,the volt-ampere characteristic and zero-sequence power direction of the zero-sequence voltage and current in the sound sections and the fault sections are different.SLGF is random and is affected by many factors,and the accuracy of fault-section location method using a single SLGF characteristic is not high.Therefore,a method of fault-section location without setting fix threshold is proposed in this paper,which is based on multi-transient characteristic cluster.Firstly,when SLGF occurs,the volt-ampere characteristic eigenvector and the zero-sequence power direction eigenvector between the first half period of the transient zero-sequence voltage derivative and the transient zero-sequence current of each section are extracted.Then,the two eigenvectors are normalized and constructed synthetic eigenvector.Finally,the sound sections and the fault sections are distinguished by the fuzzy C-means clustering algorithm without setting fix threshold.The experimental results verified that the proposed method could work effectively.

关 键 词：配电网 单相接地故障 区段定位 伏安特性 零序功率方向 模糊C均值聚类 

分 类 号：TM862[电气工程—高电压与绝缘技术]