基于先验知识的机载雷达自适应训练样本选取的方法  

Prior knowledge-based adaptive training sample selection method for airborne radar

作　　者：李虎 谢文冲[1] 熊元燚 侯铭 LI Hu;XIEWenchong;XIONG Yuanyi;HOU Ming(Air Force EarlyWarning Academy,Wuhan 430019,China)

机构地区：[1]空军预警学院,武汉430019

出　　处：《空天预警研究学报》2023年第2期89-93,104,共6页JOURNAL OF AIR & SPACE EARLY WARNING RESEARCH

基　　金：国防科技卓越青年科学基金项目(2019-JCJQ-ZQ-006)。

摘　　要：针对实际复杂环境下机载雷达面临的均匀样本数量严重不足的问题,提出了一种基于先验知识的机载雷达自适应训练样本选取方法;针对选取得到的训练样本距离间隔较远导致明显功率差异的问题,本文方法对训练样本进行了进一步修正.该方法从距离-多普勒单元角度出发,在考虑地形遮蔽影响的同时,基于地貌类型和地貌坡度构建了各训练单元的加权欧氏距离测度,实现了对均匀训练样本的有效选取.仿真结果表明,相对于现有KA-STAP方法,基于本文所提样本选取方法的STAP处理后的杂波抑制性能得到显著改善.In order to solve the problem of seriously insufficient uniform sample quantity of airborne radar in the actual complex environment,this paper proposes an adaptive training sample selection method for airborne radar based on prior knowledge.Aiming at the problem that the selected training samples have obvious power differences due to the far distance interval,the proposed method further modifies the training sample.Starting from the perspective of range-Doppler element,while considering the influence of terrain occlusion,the paper constructs the weighted Euclidean distance measures for each training unit based on terrain type and terrain slope,achieving effective selection of uniform training samples.The simulation results show that compared with the existing KA-STAP method,the clutter suppression performance after STAP processing based on the sample selection method proposed in this paper has been significantly improved.

关 键 词：先验知识 机载雷达 训练样本选取 空时自适应处理 非均匀杂波抑制 

分 类 号：TN958[电子电信—信号与信息处理]