基于最小二乘支持向量机的改进型GIS局部放电识别方法.pdfVIP

第 35 卷 第 11 期 电 网 技 术 Vol. 35 No. 11 2011 年 11 月 Power System Technology Nov. 2011  文章编号：1000-3673 （2011 ）11-0178-05 中图分类号：TM 835 文献标志码：A 学科代码：470·4034 基于最小二乘支持向量机的 改进型 GIS 局部放电识别方法 王天健，吴振升，王晖，刘栋 （北京交通大学电气学院，北京市 海淀区 100044 ） An Improved Approach to Recognize Partial Discharge in GIS Based on Minimum Least Square-Support Vector Machine WANG Tianjian, WU Zhensheng, WANG Hui, LIU Dong (School of Electrical Engineering, Beijing Jiaotong University, Haidian District, Beijing100044, China) ABSTRACT: The approach of minimum least square-support 0 引言 vector machine (LS-SVM) is used to recognize the type of partial discharge (PD) occurred in gas insulated switchgear 气体绝缘组合电器(gas insulated switchgear，GIS) (GIS). After rapid classification of signals, the characteristic 在使用过程中经常出现 2 种类型缺陷：母线尖刺和内 parameters of spectrogram of PD characteristics based on 部绝缘子金属悬浮电极。如不及时处理这些故障会严 phase distribution is used as the foundation to recognize PD 重影响电气设备安全稳定运行，这些缺陷出现时发生 type by LS-SVM. The fast classification processing of signals 的局部放电具有不同的特征，因此通过局部放电模式 mainly includes two parts: the extraction of time-frequency [1-3] characteristic of signals and fuzzy C-means clustering, and they 识别可区分不同类型的缺陷，避免出现重大事故 。 divide the group of time-frequency points into several signal 近年来，人工神经网络智能技术，如 BP 神经网 subgroups consisting of similar signals. Results of simulation 络广泛应用于电气设备局部放电的模式识别，取得 tests