一种面对雷达信号分选的无参数快速聚类算法  

A Parameter-free Fast Clustering Algorithm for Radar Signal Sorting

作　　者：彭泽宇 束坤 PENG Zeyu;SHU Kun(The Eighth Research Academy of CSSC,Yangzhou 225101,China)

机构地区：[1]中国船舶集团有限公司第八研究院,江苏扬州225101

出　　处：《舰船电子对抗》2025年第2期52-57,共6页Shipboard Electronic Countermeasure

摘　　要：针对基于密度的噪声应用空间聚类(DBSCAN)算法在雷达信号预分选中需要人为设置聚类参数、对密度分布不均雷达信号聚类准确度低、计算复杂度高的问题,提出了一种基于粒子群算法和网格划分的无参数快速聚类(GPSO-DBSCAN)算法。该算法通过粒子群算法自适应获得DBSCAN聚类最优参数,通过网格划分和分级聚类增强了对密度分布不均雷达信号的聚类能力,并降低了计算复杂度,实现了准确、快速聚类。仿真结果表明,该算法能够自适应、准确快速完成密度分布不均雷达信号的聚类。To address the problems of density-based spatial clustering of applications with noise(DBSCAN)algorithm in radar signal pre-sorting,including manual parameter setting requirements,low clustering accuracy for uneven-density radar signals,and high computational complexity,this paper proposes a parameter-free fast clustering algorithm based on grid-based particle swarm optimization DBSCAN(GPSO-DBSCAN).The algorithm adaptively acquires optimal DBSCAN clustering parameters through particle swarm optimization,enhances clustering capability for uneven-density radar signals via grid partitioning and hierarchical clustering,and reduces computational complexity to achieve accurate and rapid clustering.Simulation results demonstrate that the proposed algorithm can complete clustering tasks for radar signals with uneven density distribution adaptively,accurately and efficiently.

关 键 词：雷达信号分选 基于密度的噪声应用空间聚类算法 无参数 粒子群算法 网格单元 

分 类 号：TN971.1[电子电信—信号与信息处理]