基于融合算法的多参数雷达信号分选  

Multi-parameter Radar Signal Sorting Based on Fusion Algorithm

作　　者：陈恒 张友益[2] 王玉梅 CHEN Heng;ZHANG Youyi;WANG Yumei(Jiangsu University of Science and Technology,Zhenjiang 212001;No.723 Institute of CSIC,Yangzhou 225001)

机构地区：[1]江苏科技大学,镇江212001 [2]中国船舶重工集团公司七二三研究所,扬州225001

出　　处：《舰船电子工程》2018年第9期66-70,共5页Ship Electronic Engineering

摘　　要：针对传统单一分类器在雷达信号分选中分选准确率差的问题,论文提出了利用自组织特征映射神经(SOFM)网络结合k-means算法对常规三参数(脉宽PW、到达角DOA、载频RF)进行分选。该方法主要采用先分后合的思想,意在将信号参数分成多个模块,然后对这些模块进行再组合,解决了部分参数交叠严重的问题。对同样的雷达参数,分别利用这两种单一算法和融合算法进行分选,仿真结果表明,融合算法取得了更好的分选效果。In order to solve the problem of poor accuracy of single classifier in radar signal sorting,this paper proposes a new method to estimate the accuracy of three-parameters(pulse width PW,angle of arrival DOA,Carrier frequency RF)for sorting.This method mainly uses the idea of dividing points first and then combines,divides the signal parameters into multiple modules,and then recombines these modules to solve the problem that some parameters overlap seriously.For the same radar parameters,the two single algorithms and the fusion algorithm are respectively used for sorting.The simulation results show that the fusion algorithm achieves has achieved sorting results.

关 键 词：SOFM网络 K-MEANS算法 信号分选 

分 类 号：TN911.6[电子电信—通信与信息系统]