基于强化学习的雷达自适应波形调度杂波抑制方法  

Radar adaptive waveform scheduling method for clutter suppression based on reinforcement learning

作　　者：崔炜程[1] 冯阳 郭国强[1] CUI Weicheng;FENG Yang;GUO Guoqiang(Nanjing Research Institute of Electronics Technology,Nanjing 210039,China)

机构地区：[1]南京电子技术研究所,南京210039

出　　处：《空天预警研究学报》2024年第4期258-263,共6页JOURNAL OF AIR & SPACE EARLY WARNING RESEARCH

摘　　要：面向地面及舰载等预警雷达广域作战环境下杂波抑制问题,传统单一的调度方式难以满足高性能探测需求.为此在杂波环境自适应感知基础上,面向雷达杂波抑制任务,提出一种基于强化学习的雷达自适应波形调度杂波抑制方法.通过杂波剩余和资源消耗等维度构建雷达波形综合奖励函数,利用Q-学习和DQN强化学习模型进行雷达策略训练,实现雷达自适应波形参数调整.仿真结果表明,基于强化学习的自适应波形调度方法可以有效利用雷达系统资源,提升复杂杂波环境下雷达探测性能.Facing the problem of clutter suppression in the wide-area combat environment of ground and shipborne early warning radar,the traditional single scheduling mode has difficulty meeting the demand of high-performance detection.To this end,on the basis of adaptive perception of clutter environment,a radar adaptive waveform scheduling clutter suppression method based on reinforcement learning is proposed for radar clutter suppression task.And then,the integrated reward function of radar waveform is constructed through the dimensions such as clutter residual and resource consumption.Finally,Q-learning and DQN reinforcement learning models are used for radar strategy training,with radar adaptive waveform parameter adjustment achieved.Simulation results show that the adaptive waveform scheduling method based on reinforcement learning can effectively utilize radar system resources and improve radar detection performance in complex clutter environment.

关 键 词：雷达杂波抑制 强化学习 自适应波形调度 环境感知 

分 类 号：TN957[电子电信—信号与信息处理]