基于FCM算法的高压开关柜局部放电状态评价技术  

Partial Discharge State Evaluation Technology for High-voltage Switchgear Based on FCM Algorithm

作　　者：罗颖强 冯欢 胡海淼 陈州浩 LUO Yingqiang;FENG Huan;HU Haimiao;CHEN Zhouhao

机构地区：[1]国网浙江省电力有限公司磐安县供电公司,浙江金华322300

出　　处：《电力系统装备》2025年第2期50-51,93,共3页Electric Power System Equipment

摘　　要：文章探讨了基于FCM算法的高压开关柜局部放电状态评价技术,选取了5个开关柜状态量(TEV幅值、TEV脉冲数、超声波幅值、设备信息评分、预试定检数据评分),并分析了这些状态量所表征的信息。针对变电站背景噪声干扰问题,提出了一种基于检测值突变量的算法,再通过分析局部放电信号的突变量,识别并修正背景噪声干扰导致的异常信号。介绍了开关柜状态评价的模糊聚类算法,包括数据集构建、初始聚类中心的设定、初始隶属度矩阵的生成、隶属度函数的计算、聚类中心的更新与修正、目标函数的计算及收敛条件的判断。通过上述步骤,该算法能够有效抑制背景噪声干扰,提高局部放电检测的准确性,为高压开关柜的状态评价提供可靠的技术支持。This article explores the partial discharge state evaluation technology of high-voltage switchgear based on FCM algorithm.Five switchgear state variables(TEV amplitude,TEV pulse number,ultrasonic amplitude,equipment information score,and pre test calibration data score)are selected,and the information represented by these state variables is analyzed.Subsequently,an algorithm based on detection value mutation was proposed to address the issue of background noise interference in substations.By analyzing the mutation of partial discharge signals,abnormal signals caused by background noise interference were identified and corrected.Finally,the fuzzy clustering algorithm for evaluating the status of switchgear was introduced in detail,including dataset construction,setting of initial clustering centers,generation of initial membership matrices,calculation of membership functions,update and correction of clustering centers,calculation of objective functions,and determination of convergence conditions.Through the above steps,the algorithm can effectively suppress background noise interference,improve the accuracy of partial discharge detection,and provide reliable technical support for the state evaluation of high-voltage switchgear.

关 键 词：FCM算法 高压开关柜 局部放电 状态评价技术 

分 类 号：TM591[电气工程—电器]