气体绝缘组合电器多局部放电源的检测与识别

介绍了基于宽带检测的气体绝缘组合电器(gas insulated switchgear，GIS)多局放源检测与识别技术。该技术由脉冲群快速分类和基于最小二乘支持向量机(least square-support vector machine，LS-SVM)的基于相位分布的局部放电谱图(phase resolved partial discharge，PRPD)识别2个主要模块组成。其中脉冲群快速分类模块由基于脉冲波形的时频参数提取和竞争学习网络无监督聚类实现，它将脉冲群分为若干个由相同波形特征脉冲组成的子脉冲群。PRPD放电谱图识别模块对各子脉冲群对应的PRPD谱图进行放电指纹提取，并使用GIS绝缘缺陷特征数据库训练得到的LS-SVM判别函数对各子脉冲群进行识别。仿真和试验结果均验证了该技术的可行性和实用性。

Detection and Identification Techniques for Multi-PD Source in GIS

A novel detection and identification approach of multi-partial discharge (PD) in gas insulated switchgear (GIS) based on wideband detection technique is presented in this paper. The realization of the approach involves the classification of PD pulse sequence and identification of each sub-group with phase resolved partial discharge (PRPD) based on least squares support vector machine (LS-SVM). The classification module is composed of the time-frequency feature extraction for PD pulses and the competitive learning network (CLN) unsupervisory clustering, which divides PD pulses into sub-groups, in each group PD pulses show homogeneous stochastic features. Then the fingerprints of each sub-group are deduced from PRPD, and the identification is performed with LS-SVM discriminant function trained with the PD fingerprints produced by single insulation defect of GIS. Simulation test results show that the technique is feasible and practical, which provides a good approach for developing the multi-PD detection and identification system for GIS.