单脉冲雷达多目标分辨的遗传算法实现  

Discrimination of Multiple Targets with Monopules Radar via Genetic Algorithm

作　　者：朱玉鹏[1] 李朝伟[1] 庄钊文[1] 

机构地区：[1]国防科技大学电子科学与工程学院ATR实验室,长沙410073

出　　处：《现代雷达》2005年第2期29-32,共4页Modern Radar

基　　金：国防预研项目(编号: 413010701-2)

摘　　要：常规单脉冲雷达一般对落在同一距离单元、同一波束内的多个目标难以分辨。提出一种基于单脉冲雷达进行多目标分辨的方法。该方法利用雷达回波的和差信号经过同相正交解调获得多维观测数据,建立关于目标位置和数目参数的似然比函数,采用改进的遗传算法对多目标参数进行估计,并基于最小描述长度准则进行目标数确定。在遗传算法中引入了精英准则、局部寻优、实数表示等多种策略,有效提高了多目标参数搜索性能。仿真实验证明了该算法的整体有效性。Multiple targets located in one resolution cell can not be discriminated with ordinary monopulse radar. A multiple target discrimination method with monopulse radar is suggested in this paper. Likelihood function about the multiple targets' location and number parameter is established basing on the multi dimensional observation data which are given by the in phase quadrature modulation of radar echo. Then the multiple targets' parameters are estimated through modified genetic algorithm. A minimum description length (MDL) criterion is used to detect the number of targets. Elite criterion , local optimization and real representation are introduced here. Simulation result shows that this method is effective.

关 键 词：单脉冲雷达 分辨单元 遗传算法 最小描述长度 

分 类 号：TN958.4[电子电信—信号与信息处理]