首页 | 本学科首页   官方微博 | 高级检索  
     

基于鸟鸣功率谱密度的输电线路涉鸟故障相关鸟种智能识别
引用本文:况燕军,邱志斌,李阳林,张宇,陈瀚翔,卢祖文. 基于鸟鸣功率谱密度的输电线路涉鸟故障相关鸟种智能识别[J]. 水电能源科学, 2021, 39(2): 156-159,155
作者姓名:况燕军  邱志斌  李阳林  张宇  陈瀚翔  卢祖文
作者单位:国网江西省电力有限公司电力科学研究院,江西南昌330096;南昌大学信息工程学院能源与电气工程系,江西南昌330031
基金项目:国网江西省电力有限公司科技项目(52182018000W);江西省青年科学基金资助项目(20192BAB216028);南昌大学研究生创新专项(CX2019070)。
摘    要:鸟类活动对架空输电线路的运行影响较大,不同鸟种引起的线路故障差异明显,为了提高涉鸟故障防治的针对性,需分类识别输电线路相关鸟种.以鸟粪类、鸟巢类、鸟体短接类、鸟啄类四种涉鸟故障对应的8种代表性鸟种为例,通过分帧加窗、端点检测等预处理,采用离散傅里叶变换和Welch算法获取不同鸟鸣信号的功率谱密度(PSD)图,从中提取1...

关 键 词:输电线路  涉鸟故障  鸟种识别  功率谱密度  随机森林

Intelligent Recognition of Bird Species Related to Transmission Line Bird Faults Based on Power Spectrum Density of Birdsong
KUANG Yan-jun,QIU Zhi-bin,LI Yang-lin,ZHANG Yu,CHEN Han-xiang,LU Zu-wen. Intelligent Recognition of Bird Species Related to Transmission Line Bird Faults Based on Power Spectrum Density of Birdsong[J]. International Journal Hydroelectric Energy, 2021, 39(2): 156-159,155
Authors:KUANG Yan-jun  QIU Zhi-bin  LI Yang-lin  ZHANG Yu  CHEN Han-xiang  LU Zu-wen
Affiliation:(Elecrie Power Research Institute of State Grid Jiangxi Elctric Power Company,Nanchang 330096,China;Department of Energy and Electrical Engineering,Information Engineering School,Nanchang University,Nanchang 330031,China)
Abstract:Bird activities severely affect the safe operation of overhead transmission lines.Line faults caused by differ-ent bird species varies significantly.In order to improve the pertinence of bird-related outage prevention,it is necessary to conduct classification and recognition of the birds related to transmission lines.In this paper,8 kinds of representative bird species were taken for study,respectively corresponding to four types of bird-related outages.including bird excre-ment outage,nesting outage,bird short-circuit outage and pecker outage.The birdsong signals were preprocessed by framing,windowing,and endpoint detection.And then the power spectrum density(PSD)figures were obtained by dis-crete Fourier trans-formation(DEF)and Welch algorithm.The PSD features of 129 frequency points were extracted.A classification recognition model was established by random forest(RF)and the PSD features were taken as input parame-ters.The model training and classification testing of different bird species were conducted.The recog nition accuracy of 8 bird species varies from 83.3%to 100%.This study improves the pertinence of prevention countermeasures for bird relat-ed outages.
Keywords:transmission line  bird related outages  bird species recognition  power spectrum density  random forest
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号