采用小波分析与神经网络技术的局部放电统计识别方法.pdfVIP

第22 卷 第9 期 中 国 电 机 工 程 学 报 Vol.22 No.9 Sep. 2002 2002 年9 月 Proceedings of the CSEE ©2002 Chin.Soc.for Elec.Eng. 文章编号 0258-8013 (2002) 09-0001-06 采用小波分析与神经网络技术的 局部放电统计识别方法 淡文刚, 陈祥训, 郑健超 ( 中国电力科学研究院 100085 北京) CLASSIFICATION OF PARTIAL DISCHARGE DISTRIBUTION PATTERNS USING WAVELET TRANSFORM AND NEURAL NETWORK DAN Wen-gang, CHEN Xiang-xu , ZHENG Jian-chao (China Electrical Power Research Institute, Beijing 100085, China) ABSTRCT: Wavelet transform(WT) can be used for detecting 以新定义的特征量组成输入矢量 用基于 BP 算法的前馈 and recognizing parital discharges (PDs). Two key points for 型神经网络 对6 种典型PDs 及加上4 种混合PDs 组成的 this purpose are choosing right wavelet and extracting proper 共 10 种PDs 进行了识别测试 识别效果远远优于现有方法 features from the WT coefficeinces. 144 sets of PD pulse data 水平 are obtained from 6 types of insulation defect models. Using the 关键词 局部放电模式识别 小波变换 特征提取 神经 Cardinal B-spline derivative wavelets all the PDs are picked up 网络 successfully from strong EM interference background encountered in the measurement field, and 2D and 3D statistical 中图分类号 O177.6 ; TN911.4 文献标识码 A distribution patterns of these PDs are constructed. Then new 1 引言 WT ways to decompose distribution patterns are presented, new features in WT-domain to identify PDs are defined, and the 近年来 小波变换技术在电力系统中的应用 feature extraction methods are developed. Taking these features 研究日渐广泛 内容涉及电能质量监测 电力系统 as fingerprint parameters 10 types of distribution patterns (6 扰动分析 电磁暂态识别 变压器涌流识别 配电 original SPDs and 4 mixed SPDs) are recognized correctly by a 线故障