首页 | 本学科首页   官方微博 | 高级检索  
     检索      

一种基于奇异值分解的舰船辐射噪声目标识别算法
引用本文:潘秀琴,张春华,黄海宁,张洪.一种基于奇异值分解的舰船辐射噪声目标识别算法[J].应用声学,2005,24(2):108-113.
作者姓名:潘秀琴  张春华  黄海宁  张洪
作者单位:1. 中国科学院声学研究所,北京,100080
2. 成都电子科技大学,成都,610045
摘    要:本文针对舰船辐射噪声目标识别问题,提出了一种基于奇异值分解(SVD)的辐射噪声爿标识 别算法。该算法充分利用了不等权值SVD的空间滤波作用,来消除辐射噪声中复杂的干扰成分;结 合二阶累积量谱进行特征分析及提取;然后根据所要解决的实际问题,设计了合适的概率密度函数, 并对其进行训练,作为识别模板;进而根据距离分类准则设计了分类函数,以完成识别分类。运用实 际的舰船辐射噪声数据,进行了仿真实验,结果表明了本文算法的可行性和有效性。

关 键 词:奇异值分解  二阶累积量(谱)  概率密度  识别

An algorithm of object recognition of ship noise based on SVD
PAN Xiu-Qin,ZHANG Chun-Hu,HUANG Hai-Ning and ZHANG Hong.An algorithm of object recognition of ship noise based on SVD[J].Applied Acoustics,2005,24(2):108-113.
Authors:PAN Xiu-Qin  ZHANG Chun-Hu  HUANG Hai-Ning and ZHANG Hong
Institution:Institute of Acoustics, Chinese Academy of Sciences, Beijing 100080;Institute of Acoustics, Chinese Academy of Sciences, Beijing 100080;Institute of Acoustics, Chinese Academy of Sciences, Beijing 100080;College of E.E, UESTC, Chengdu 610045
Abstract:An algorithm of object recognition for noise from ship is presented. In the algorithm, SVD with un-equivalent weight is used to cancel the complicated disturbance, and the second cumulant spectrum is adopted for feature analysis and extraction, then recognition template based on probability density is designed and trained, and the classification function is given according to distance sorting rules. Simulation on the actual noises from ships is carried out, and the results illustrate that the algorithm is effective and valid.
Keywords:SVD  The second cumulant (spectrum)  Probability density  Recognition
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《应用声学》浏览原始摘要信息
点击此处可从《应用声学》下载免费的PDF全文
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