基于多尺度形态分解谱熵的雷击干扰与短路故障识别研究  

Study of Lightning Interference and Fault Identification Based on Multi-Scale Morphological Decomposition Spectral Entropy

作　　者：陈仕龙[1] 曹蕊蕊[1] 毕贵红[1] 李兴旺[1] 荣俊香 

机构地区：[1]昆明理工大学电力工程学院,云南昆明650500

出　　处：《昆明理工大学学报（自然科学版）》2014年第6期74-81,共8页Journal of Kunming University of Science and Technology(Natural Science)

基　　金：国家自然科学基金项目(51267008);昆明理工大学校人培基金项目(kkz3201304019)

摘　　要：研究和识别输电线路直击雷的暂态特征对研制基于暂态量的保护及提高可靠性都具有十分重要的意义.针对雷击未故障、雷击故障和接地短路故障,提出了基于多尺度形态分解谱熵的雷击干扰识别方法.对不同的故障类型进行多尺度分解,分别计算在不同尺度上的能谱熵和奇异谱熵,作为待识别特征矢量.建立标准故障模式矩阵,对待识别样本的特征矢量与标准故障模式进行灰色关联分析,根据关联度的大小,判断待识别样本信号属于哪种故障类型.大量仿真计算验证以上所提方法正确、有效.Study on transient characteristics of overhead transmission lines due to direct lightning stroke is signif-icant to the development of transient based protection and its reliability improvement.A method for non-fault lightning stroke,fault lightning stroke and line short circuit fault based on grey relational analysis and multi -scale morphological decomposition spectral entropy is proposed.The multi-scale morphological decomposition is performed for the different fault types,whose power spectral entropy and singular spectral entropy in different scale domains are computed.The two indicators are taken as character-vector.Then,the standard fault mode matrix is built.The grey relational analysis between the character-vector and standard fault mode matrix of the sample signal to be detected is conducted to distinguish the different fault types according to the value of grey rel-evancy.The simulations demonstrate the correctness and effectiveness of the above proposed approach.

关 键 词：直击雷 灰色关联分析 能谱熵 奇异谱熵 标准故障模式矩阵 

分 类 号：TM713[电气工程—电力系统及自动化]