基于宽深融合神经网络的雷达HRRP目标识别  

Radar HRRP Target Recognition Based on Broad-Deep Fusion Neural Networks

作　　者：李显丰 LI Xianfeng(Xi’an Electronic Engineering Research Institute,Xi’an 710100)

机构地区：[1]西安电子工程研究所,西安710100

出　　处：《火控雷达技术》2023年第3期16-24,共9页Fire Control Radar Technology

摘　　要：目标识别是现代雷达的一个十分重要的发展方向。使用高距离分辨率雷达可获得目标的一维距离像,其中包含很多可用于目标识别的信息。本文针对一维距离像数据,基于宽深融合神经网络对雷达目标识别进行了研究。首先,探讨了宽度学习这一新概念,它拥有简单的网络结构,横向扩展和快速训练的特点。然后,本文开创性地提出了三种宽深融合神经网络,兼具了训练速度快和准确率高的特点,实验表明宽深融合神经网络有一定的运用于雷达目标识别的可行性。Target recognition is a very important development direction of modern radar.One-dimensional range profiles of the targets can be obtained by using high range resolution radar,which contain a lot of information that can be used for target recognition.In this paper,radar target recognition based on one-dimensional range profile data is studied using broad-deep fusion neural networks.Firstly,the new concept of broad learning is discussed,which features simple network structure,lateral expansion,and rapid training.Then,this paper puts forward three types of broad-deep fusion neural networks,which have the characteristics of high training speed and high accuracy.The test results show that the broad-deep fusion neural networks are feasible for radar target recognition.

关 键 词：目标识别 一维距离像 深度学习 宽度学习 宽深融合神经网络 

分 类 号：TN95[电子电信—信号与信息处理]