电工技术学报  2020, Vol. 35 Issue (23): 5015-5024    DOI: 10.19595/j.cnki.1000-6753.tces.191400
高电压与绝缘 |
基于改进K-近邻算法的XLPE电缆气隙放电发展阶段识别
陈曦, 骆高超, 曹杰, 毕茂强, 江天炎
重庆理工大学电气与电子工程学院 重庆 400054
Development Stage Identification of XLPE Cable Air-Gap Discharge Based on Improved K-Nearest Neighbor Algorithm
Chen Xi, Luo Gaochao, Cao Jie, Bi Maoqiang, Jiang Tianyan
College of Electrical and Electronic Engineering Chongqing University of Technology Chongqing 400054 China
全文: PDF (1442 KB)   HTML
输出: BibTeX | EndNote (RIS)      
摘要 对运行工况下的XLPE电缆气隙放电发展阶段进行准确识别,有利于将电缆故障扼杀在萌芽状态,保障电力系统正常运行。该文首先介绍了模拟XLPE电缆气隙放电的试验平台搭建、缺陷模型制作和特征量提取及降维的方法及步骤,基于试验观察和对大量数据样本进行聚类分析,将XLPE电缆气隙放电发展过程分为四个阶段,针对以往XLPE电缆气隙放电阶段识别模型的训练周期长、计算复杂度高和收敛速度慢等问题,该文提出一种经高斯函数加权的改进K-近邻(KNN)分类算法应用于XLPE电缆气隙放电阶段识别。对气隙放电的随机测试样本采用基于二叉树的核函数支持向量机、未改进的K-近邻算法和改进后的K-近邻算法三种算法分别进行了阶段识别。试验结果表明,改进后的K-近邻算法识别正确率高、速度快,处理含噪信号鲁棒性好,相比另两种算法更优。
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
陈曦
骆高超
曹杰
毕茂强
江天炎
关键词 XLPE气隙放电特征参量改进K-邻近算法支持向量机    
Abstract:It is essential to identify development stages of air-gap discharge in XLPE cable accurately under operation. It will not only help to prevent the cable fault, but to safeguard operation of power system. In this paper, the test platform and artificial defect sample manufacture method were introduced firstly. Then feature extraction and dimension reduction steps on a large number of experimental samples were described. Based on experimental observation and cluster analysis, the development process of air-gap discharge was classified into four stages accordingly. Aiming at the problems such as long training cycle, high computational complexity and slow convergence rate, this paper proposes an improved K-nearest neighbor (KNN) r classification algorithm weighted by Gaussian function to identify the air-gap discharge stage of XLPE cable. Three kinds of algorithms, namely kernel support vector machine (SVM) based on binary tree, unimproved KNN and improved KNN, were used to identify the random test samples of air gap discharge. Results show that improved KNN algorithm has high recognition accuracy, fast recognition speed and good robustness in processing noisy signals, which is better than the other two algorithms.
Key wordsXLPE    air-gap discharge    characteristic parameter    improved K-nearest neighbor (KNN)    support vector machine   
收稿日期: 2019-11-01     
PACS: TM85  
基金资助:重庆市基础研究与前沿探索项目(重庆市自然科学基金)(cstc2018jcyjAX0295)、重庆市教委科学技术研究项目(KJQN202001146)和国家自然科学基金(51607019)资助项目
通讯作者: 骆高超 男,1994年生,硕士研究生,研究方向为电力设备在线监测与智能诊断技术。E-mail:472473953@qq.com   
作者简介: 陈 曦 男,1986年生,博士,讲师,研究方向为电力设备在线监测与智能诊断技术、能源互联网、能源经济与市场等。E-mail:chenxi1986@cqut.edu.cn
引用本文:   
陈曦, 骆高超, 曹杰, 毕茂强, 江天炎. 基于改进K-近邻算法的XLPE电缆气隙放电发展阶段识别[J]. 电工技术学报, 2020, 35(23): 5015-5024. Chen Xi, Luo Gaochao, Cao Jie, Bi Maoqiang, Jiang Tianyan. Development Stage Identification of XLPE Cable Air-Gap Discharge Based on Improved K-Nearest Neighbor Algorithm. Transactions of China Electrotechnical Society, 2020, 35(23): 5015-5024.
链接本文:  
https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.191400          https://dgjsxb.ces-transaction.com/CN/Y2020/V35/I23/5015