融合核主成分与随机森林的瓷绝缘子污秽等级诊断方法研究  

Research on Fusion of Kernel Principle Component and Random Forest for Determination of Porcelain Insulators Pollution Class

作　　者：刁培松 DIAO Peisong(Shandong Electric Power Company Haiyang Power Supply Company,Haiyang 265100,China)

机构地区：[1]山东省电力公司海阳供电公司

出　　处：《电瓷避雷器》2019年第5期188-193,共6页Insulators and Surge Arresters

摘　　要：由瓷绝缘子表面污秽引起的闪络放电严重影响了电力系统的安全稳定运行,针对传统分析方法无法对瓷绝缘子污秽程度特征量的不明显和非线性属性进行准确反映,且瓷绝缘子污秽等级诊断准确率低的现状,提出了一种融合核主成分与随机森林的瓷绝缘子污秽等级诊断新方法。在实验室模拟信号和现场实测信号的基础上,通过该诊断新方法与其它传统方法诊断结果的对比分析,结果表明该方法具有更高的准确率和更好的噪声容忍度,能够有效地对瓷绝缘子的污秽等级进行诊断,可为瓷绝缘子的污秽监测及污闪的预防提供有效借鉴和参考。The flashover discharge caused by porcelain insulator surface contamination seriously affects the safe and stable operation of power system.In view of the current situation that traditional method cannot accurately reflect the non-obvious and non-linear attributes of the pollution degree characteristics,and the low accuracy of porcelain insulator pollution level diagnosis,a new method fused kernel principal component with random forest for porcelain insulator pollution class diagnosis is proposed.Based on the laboratory simulation test and measured signal for on-site operation,and through the comparative analysis of new diagnosis method with other traditional methods,the result shows that the diagnosis method proposed in this paper have higher accuracy and better noise tolerance,and can diagnosis the porcelain insulator pollution class effectively.It can provide effective reference for the monitoring of the porcelain insulator pollution and the prevention of pollution flashover.

关 键 词：瓷绝缘子 污秽等级 核主成分 随机森林 泄漏电流 

分 类 号：TM2[一般工业技术—材料科学与工程]