基于非线性加权K-相关的核爆地震识别

详细信息 本馆镜像全文    |  推荐本文 | | 获取馆网全文

英文篇名：Identification of underground nuclear explosions and natural earthquakes based on K-correlation of nonlinear weighting 中文刊名：核电子学与探测技术 英文刊名：Nuclear Electronics & Detection Technology 作者：刘代志 ; 李夕海 ; 何志文 ; 张斌 英文作者：LIU Dai-zhi ; LI Xi-hai ; HE Zhi-wen ; Zhang Bin (The Second Artillery Institute of Engineering Xi'an of Shaanxi Prov. 710025 ; China) 中文关键词：特征相关 ; 非线性加权 ; 核爆炸 ; 地震 英文关键词：feature correlation ; nonlinear weighting ; nuclear explosion ; natural earthquake 出版日期：2005-11-20 机构：第二炮兵工程学院602室,第二炮兵工程学院602室,第二炮兵工程学院602室,第二炮兵工程学院602室 陕西西安 710025 ,陕西西安 710025 ,陕西西安 710025 ,陕西西安 710025 年：2005 期：06 出版单位：核电子学与探测技术

摘要

地下核爆炸和天然地震的识别一直是国际核查的重点研究内容之一,基于地下核爆炸和天然地震事件的相关性,提出了一种非线性加权K-相关分类方法,对地下核爆炸和天然地震进行分类,详细阐述了它的分类原理,讨论了训练样本个数和K值两个参数的选取问题,并将它和最大相关法及K-近邻法进行了比较。实际分类结果表明了该方法的有效性。
Automatic discrimination of underground nuclear explosions and natural earthquakes is al ways worth to study because of the problem of classification accuracy. In this paper, a method of non linear weighting .K-correlation is proposed to discriminate the underground nuclear explosions and natu ral earthquakes, the basic idea is that the sum of K correlation coefficients for each class is used as a criterion to discriminate these two kinds of events, and to magnify the difference of this criterion to i dentify these two events with higher probability, a method of nonlinear weighting is used. Besides, the comparison with the maximum correlation classification and k-nearest neighbor classification is expatiat ed. Theoretical analysis and practical identification results indicate this method is feasible.

引文

[1]CD/1341,Progress report to the conference on dis armament on the forty-first session of the ad hoc group of scientific experts to consider international cooperative measures to detect and identify seismic events, Conference on Disarmament, 1993.
    [2]吉小军,等．相关分析在特征选择中的应用[J]．华 北工学院测试技术学报,2001,15(1):15．
    [3]张善文,等．基于多重变换的一种雷达目标识别方 法[J]红外与毫米波学报,2002,21(4):318．
    [4]梁路宏,等．基于模板匹配与支持矢量机的人脸检 测[J]．计算机学报,2002,25(1):22．
    [5]朱永松,等．基于相关系数的相关匹配算法的研究 [J]．信号处理,2003,19(6):531．
    [6]刘代志．吸引子分析与核爆地震模式识别[J]．电子 学报,1997,25(10):122
    [7]邹红星．地下核爆炸和天然地震自动识别研究[D]． 西安:第二炮兵工程学院,1995．
    [8]Fukunaga K, et al. Optimization of K-nearest neighbor density estimates[J]. IEEE Trans Infor mation Theory, 1973, 19:320.