雷达与对抗2024,Vol.44Issue(1) :21-25.DOI:10.19341/j.cnki.1009-0401.2024.01.005

基于神经网络的雷达辐射源分类方法

Radar emitter classification method based on neural network

蓝天亮 茅玉龙 杨明远
雷达与对抗2024,Vol.44Issue(1) :21-25.DOI:10.19341/j.cnki.1009-0401.2024.01.005

基于神经网络的雷达辐射源分类方法

Radar emitter classification method based on neural network

蓝天亮 1茅玉龙 1杨明远1
扫码查看

作者信息

  • 1. 中国船舶集团有限公司第八研究院,南京 211153
  • 折叠

摘要

针对现代雷达对目标检测和识别的要求,结合人工智能设计一种神经网络模型.首先提取雷达脉冲信号的多维特征,然后将特征转化为图片输送到模型中进行训练,最后对信号进行分类.使用实际采集的导航雷达数据进行实验验证,验证结果表明该模型对已知的导航雷达辐射源脉冲信号分类正确率达到99.21%,也能在一定程度上区分未知雷达信号,表明神经网络模型具有较强的分类识别能力.

Abstract

In order to meet the requirements of modern radars for target detection and recognition,a neural network model is designed based on artificial intelligence,which first extracts the multi-di-mensional features of the radar pulse signals,then converts the features into pictures and sends them to the model for training,and finally classifies the signals.Based on the actual collected navi-gation radar data,the test results show that the model is able to classify the pulse signals of the known navigation radar emitters with an accuracy of 99.21%,and it can also distinguish unknown radar signals to a certain extent,indicating that the neural network model has strong classification and recognition capability.

关键词

多维特征/雷达分类/神经网络

Key words

multi-dimensional features/radar classification/neural network

引用本文复制引用

出版年

2024
雷达与对抗
南京船舶雷达研究所

雷达与对抗

影响因子:0.158
ISSN:1009-0401
参考文献量3
段落导航相关论文