基于多帧聚类的紧凑型HFSWR虚假点迹识别方法  

A false plot identification method based on multi-frame clustering for compact HFSWR

作　　者：孙伟峰[1] 赵林林 纪永刚 戴永寿[1] SUN Weifeng;ZHAO Linlin;JI Yonggang;DAI Yongshou(College of Oceanography and Space Informatics,China University of Petroleum(East China),Qingdao 266580,China)

机构地区：[1]中国石油大学(华东)海洋与空间信息学院,山东青岛266580

出　　处：《系统工程与电子技术》2024年第2期419-427,共9页Systems Engineering and Electronics

基　　金：国家自然科学基金(62071493,61831010)资助课题。

摘　　要：紧凑型高频地波雷达发射功率低,目标检测时信噪比低、虚警率高,会产生大量虚假点迹,影响后续目标跟踪性能。为了滤除虚假点迹,利用目标的运动特性,提出了一种多帧聚类与极限学习机分类两级级联的虚假点迹识别方法。首先,利用基于最优邻域尺寸的多帧聚类方法,将连续多帧中与待识别点迹属于同一潜在目标的点迹聚类成簇。然后,计算簇内待识别点迹与其相邻帧内点迹的距离-多普勒速度的差分值,以其为特征利用极限学习机辨识虚假点迹。实验结果表明,所提方法能够准确将属于同一目标的点迹聚类,虚假点迹识别率达到95%。Compact high-frequency surface wave radar(HFSWR)has low signal-to-noise ratio and high false alarm rate in target detection due to its low transmit power,a large number of false plots will be produced,which degrades the target tracking performance.In order to remove the false plots,a two-stage cascaded false plot identification method including multi-frame clustering module and extreme learning machine based classification module is proposed with target motion characteristics well explored.Firstly,the multi-frame plot clustering method based on optimal neighborhood size is utilized to cluster the potential plots belonging to the same target with the plot to be identified in consecutive multiple frames.Then,the differences in terms of range-Doppler velocity between the plot to be identified and plots in its neighbor frames are calculated as features,and the extreme learning machine is applied to these features to identify the false plots.Experimental results demonstrate that the proposed method can cluster the plots belonging to the same target accurately,and achieves a false plot identification rate of 95%.

关 键 词：紧凑型地波雷达 虚假点迹识别 多帧聚类 极限学习机 

分 类 号：TN953[电子电信—信号与信息处理]