东北电力技术2025，Vol.46Issue(9)：22-26,5.
基于局部放电信号时域特征的电缆故障智能识别
Intelligent Identification of Cable Faults Based on Time Domain Characteristics of Partial Discharge Signals
斯捷 1陈植 1赵建 1金利引1
作者信息
- 1. 浙江图盛输变电工程有限公司,浙江 温州 325000
- 折叠
摘要
Abstract
It proposes an intelligent recognition method based on the feature extraction of partial discharge pulse signals for addressing the challenge of operational reliability assessment in the maintenance and testing of distribution cable lines.By employing high-speed sampling hardware to perform broadband sampling of cable partial discharge current pulses,it extracts 13 multi-dimensional time-do-main features to establish a multi-dimensional time-domain feature dataset for cable faults.Subsequently,it utilizes the support vector machine(SVM)to analyze these time-domain feature data,enabling intelligent identification of two or more insulating conditions of the cable.It uses this method for fault diagnosis by extracting complete time-domain feature signals.The test results show that its diagnos-tic accuracy rate increases from 56.3%to 97.6%.It significantly reduces the issues of false alarms and misses reports in warning in-formation,enhancing the operational reliability and safety of cable lines.关键词
局部放电/时域特征信号/支持向量机Key words
partial discharge/time-domain feature signal/support vector machine分类
信息技术与安全科学引用本文复制引用
斯捷,陈植,赵建,金利引..基于局部放电信号时域特征的电缆故障智能识别[J].东北电力技术,2025,46(9):22-26,5.
