基于距离-方位二维聚类的海上编队目标距离像提取  

Range profiles extraction for maritime formation targets based on range-azimuth clustering

作　　者：罗小波[1,2] 范红旗[1] 宋志勇[1] 付强[1] 

机构地区：[1]国防科学技术大学电子科学与工程学院ATR重点实验室 [2]中国人民解放军96363部队

出　　处：《系统工程与电子技术》2013年第7期1353-1361,共9页Systems Engineering and Electronics

基　　金：武器装备预研(51301020504)资助课题

摘　　要：在观测编队目标时波束内容易出现多个目标,可能引起距离像邻近甚至混叠,难以判断距离像簇与目标之间的对应关系,导致无法准确提取单个目标的距离像信息。基于距离像目标识别的需求,在海面编队结构信息的约束下,分析了波束内多目标出现概率,定义了多目标距离像之间的4种关系。利用目标的距离和方位角信息进行二维聚类来判断距离像簇与多目标之间的对应关系,从而实现波束内多个舰船目标的距离像提取。仿真实验表明该算法能够在不同信噪比和带宽条件下很好地区分编队内不同目标的距离像,为下一步基于距离像信息的目标识别和稳定跟踪提供了重要条件。When observing maritime formation targets with a missile-borne radar,multiple range profiles are likely vicinity or even overlapped in consequence of appearance of multiple targets in the antenna beam.It is difficult to judge the corresponding relation between range profile cluster and targets,which leads to the consequences that the single target range profile cannot be accurately extracted.Based on the requirements of range profile recognition and the constraints imposed by the target formation structure information,a range profile extraction algorithm is proposed.Firstly,the probability that multiple targets exist within the same beam is analyzed and the four relationships among multiple targets range profiles are defined.Subsequently,the relationship between range profile cluster and multiple targets are judged with clustering the range and azimuth information of the surface targets in the range-azimuth space.The algorithm finally extracts the range profiles of each individual ship target within the beam according to the clustering information.Simulation results demonstrate that the algorithm performs well in distinguishing different targets range profiles under a wide range of signal-to-noise ratio（SNR） and bandwidth,which provides a solid foundation for the subsequent target recognition and tracking applications.

关 键 词：距离像提取 高分辨距离像 单脉冲测角 编队目标 

分 类 号：TN953.5[电子电信—信号与信息处理]