基于DTCWT-VAE的弹道中段目标RCS识别  

Ballistic midcourse target RCS recognition based on DTCWT-VAE

作　　者：王彩云 张慧雯 王佳宁 吴钇达 常韵 WANG Caiyun;ZHANG Huiwen;WANG Jianing;WU Yida;CHANG Yun(College of Astronautics,Nanjing University of Aeronautics and Astronautics,Nanjing 211106,China;Beijing Institute of Electronic System Engineering,Beijing 100854,China)

机构地区：[1]南京航空航天大学航天学院,江苏南京211106 [2]北京电子工程总体研究所,北京100854

出　　处：《系统工程与电子技术》2024年第7期2269-2275,共7页Systems Engineering and Electronics

基　　金：国家自然科学基金(61301211);国家留学基金(201906835017)资助课题。

摘　　要：针对弹道目标雷达信号易受环境影响、目标识别准确率低的问题,提出了一种基于双树复小波变换(dual-tree complex wavelet transform,DTCWT)和变分自编码器(variational autoencoder,VAE)的弹道目标雷达散射截面(radar cross section,RCS)识别法。首先,采用DTCWT对弹道目标RCS动态数据进行预处理,再利用VAE提取目标的隐变量特征,最后用支持向量机(support vector machine,SVM)分类器进行识别。实验结果表明,与已有方法相比,该方法具有更高的识别概率,且鲁棒性较好。Aiming at the problem that the radar signal of ballistic target is easily affected by the environment and the target recognition accuracy is low,a radar cross section(RCS)recognition method of ballistic target based on dual-tree complex wavelet transform(DTCWT)and variational autoencoder(VAE)is proposed.Firstly,the dynamic datas of ballistic target RCS are preprocessed by DTCWT.Then,the hidden variable features of target are extracted by VAE.Finally,the support vector machine(SVM)classifier is used to identify the data.Experimental results show that compared with the existing methods,the proposed method has higher recognition probability and better robustness.

关 键 词：弹道目标 目标识别 雷达散射截面 双树复小波变换 变分自编码器 

分 类 号：TN957[电子电信—信号与信息处理]