基于常规参数和脉内特征参数的雷达信号分选  

Radar Signal Sorting Based on Conventional and Intravenous Characteristic Parameters

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

机构地区：[1]江苏科技大学,镇江212001 [2]中船重工集团七二三研究所,扬州225001

出　　处：《舰船电子工程》2018年第6期58-61,75,共5页Ship Electronic Engineering

摘　　要：当雷达的常规五参数若是相互交叠或者十分相近时,常规的三参数分选法不能满足实际的需要。论文在三参数基础之上,加入脉内特征参数Wpt6(小波包特征),针对自组织特征映射神经(SOFM)网络需提前设定网络规模的不足,提出一种规模自调整的SOFM网络对四参数进行分选,经过仿真证明,加入的脉内特征参数提高了雷达信号分选准确率。The conventional three-parameter sorting method can not meet the actual needs when the conventional five parame- ters of radar are overlapped or very similar. Based on the three parameters, the in-pulse feature parameter Wpt6 （wavelet packet fea- ture is added）. In order to solve the problem that the network size needs to be set in advance for the self-organizing feature map neu- ral network （SOFM）, a self-adjusting SOFM network is proposed. The parameters are sorted, and the simulation results show that the added pulse characteristic parameters improve the accuracy of radar signal sorting.

关 键 词：小波包特征 SOFM网络 信号分选 

分 类 号：TN95[电子电信—信号与信息处理]