电气技术  2019, Vol. 20 Issue (1): 18-23    DOI:
研究与开发 |
基于KNN的配电网高阻接地故障识别
刘炳南1, 郭谋发1, 陈永往2
1. 福州大学电气工程与自动化学院,福州 350116;
2. 国网晋江供电公司,福建 泉州 362200
High resistance grounding fault identification in distribution network based on KNN
Liu Bingnan1, Guo Moufa1, Chen Yongwang2
1. College of Electrical Engineering and Automation, Fuzhou University, Fuzhou 350116;
2. State Grid Jinjiang Electric Power Supply Company, Quanzhou, Fujian 353000
全文: PDF (15321 KB)   HTML (1 KB) 
输出: BibTeX | EndNote (RIS)      
摘要 配电网受实际因素的影响,运行时接触高阻抗表面会发生高阻接地故障。发生故障时,故障点电阻较大,电压、电流突变量不明显,常规保护难以准确动作。高阻接地故障常伴随着电弧熄灭与重燃,若该故障长期存在,则将导致电气设备永久性损坏,甚至发生人身触电安全事故。本文通过对母线三相电压和零序电压进行HHT带通滤波构造时频矩阵,对各频带求取标准差作为特征向量,最后利用K最邻近算法进行故障辨识。
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
刘炳南
郭谋发
陈永往
关键词 配电网高阻接地故障HHT带通滤波标准差K最邻近算法    
Abstract:The distribution network is affected by actual factors, and high impedance ground faults occur when contact with highimpedance surfaces during operation.When high impedance grounding fault occurs, the resistance of the fault point is large, and the sudden change of voltage and current is not obvious. Conventional protection is difficult to operate accurately. High resistance grounding fault is often accompanied by arc extinguishment and reignition. If the fault exists for a long time, it will lead to permanent damage of electrical equipment and even accidents of personal electric shock. In this paper, the HHT bandpass filter is used to construct the time-frequency matrix of the three-phase voltage and zero sequence voltage of the bus, and the standard deviation is taken as the eigenvector for each frequency band. Finally, the fault identification is carried out by using the K nearest neighbor algorithm.
Key wordsMV distribution networks    HIF    HHT band-pass filter    RMS    KNN   
收稿日期: 2018-07-16      出版日期: 2019-01-15
基金资助:晋江市科技计划项目(2017C006)
作者简介: 刘炳南(1993-),男,硕士,主要研究方向为配电网及其自动化技术。
引用本文:   
刘炳南, 郭谋发, 陈永往. 基于KNN的配电网高阻接地故障识别[J]. 电气技术, 2019, 20(1): 18-23. Liu Bingnan, Guo Moufa, Chen Yongwang. High resistance grounding fault identification in distribution network based on KNN. Electrical Engineering, 2019, 20(1): 18-23.
链接本文:  
http://dqjs.cesmedia.cn/CN/Y2019/V20/I1/18