系统工程与电子技术 ›› 2023, Vol. 45 ›› Issue (6): 1642-1650.doi: 10.12305/j.issn.1001-506X.2023.06.07

• 电子技术 • 上一篇    

基于角域特征PSO的海面目标HRRP识别方法

王哲昊, 简涛, 黄晓冬, 王海鹏, 刘瑜   

  1. 海军航空大学信息融合研究所, 山东 烟台 264001
  • 收稿日期:2021-09-23 出版日期:2023-05-25 发布日期:2023-06-01
  • 通讯作者: 简涛
  • 作者简介:王哲昊 (1996—), 男, 硕士研究生, 主要研究方向为信号与信息处理
    简涛 (1980—), 男, 教授, 博士, 主要研究方向为信号与信息处理
    黄晓冬 (1975—), 男, 教授, 博士, 主要研究方向为系统仿真、计算机软件
    王海鹏 (1985—), 男, 教授, 博士, 主要研究方向为信号与信息处理
    刘瑜 (1986—), 男, 教授, 博士, 主要研究方向为信息融合

HRRP recognition method for sea surface targets based on angular domain feature PSO

Zhehao WANG, Tao JIAN, Xiaodong HUANG, Haipeng WANG, Yu LIU   

  1. Research Institute of Information Fusion, Naval Aviation University, Yantai 264001, China
  • Received:2021-09-23 Online:2023-05-25 Published:2023-06-01
  • Contact: Tao JIAN

摘要:

针对特征空间中各类海面目标特征混叠严重和高分辨距离像(high resolution range profile, HRRP)的角度特征利用率低的问题, 提出了一种基于角域特征粒子群优化(particle swarm optimization, PSO)的海面目标HRRP识别方法。该方法引入HRRP的角度信息优化特征空间, 增加特征空间的整体可分性; 利用自适应分帧算法对特征空间进行角域划分, 增加特征空间的局部可分性, 并利用PSO算法确定特征空间角域划分时最优的单帧最小样本数目, 增强方法的鲁棒性与适用性。实验结果表明, 通过将特征空间优化和区域划分进行结合, 可以有效提升多类海面目标的分类识别性能, PSO算法可以有效增强方法的抗误差性和抗噪鲁棒性。

关键词: 海面目标识别, 高分辨距离像, 特征空间优化, 自适应分帧, 粒子群优化算法

Abstract:

To solve the problems of serious aliasing of various sea surface target features in feature space and low utilization rate of angle features in high resolution range profile(HRRP), a method for recognizing sea surface target HRRP based on angle domain feature particle swarm optimization(PSO) is proposed. In this method, HRRP angle information is introduced to optimize the feature space and increase the overall separability of the feature space. In order to improve the local separability of feature space, adaptive frame segmentation algorithm is used to divide feature space into angle domains. At the same time, the PSO algorithm is used to determine the optimal minimum number of samples per frame in the angular division of feature space, which enhances the robustness and applicability of the method. Experimental results show that the combination of feature space optimization and region division can effectively improve the classification and recognition performance of multi-class sea surface targets. PSO algorithm can effectively enhance the anti-error and anti-noise robustness of the method.

Key words: sea surface target recognition, high resolution range profile(HRRP), feature space optimization, adaptive frame segmentation, particle swarm optimization(PSO) algorithm

中图分类号: