基于DCN-SVA的雷达平面成像旁瓣抑制  

Sidelobe Suppression in Radar Planar Imaging based on DCN-SVA

作　　者：安泽萍 张智鸿 马世霞 胡敏 AN Ze-pinggl;ZHANG Zhi-hongg;MA Shi-xia;HU Min(Tech Traffic Engineering Group Co.Ltd.,Beijing 100083,China;China Highway Engineering Consulting Corporation,Beijing 100089,China;Research and Development Center of Transport Industry of Self-driving Technology,Beijing 100083;Research and Development Center on Intelligent Transportation CCCC,Beijing 100083,China;School of Aerospace Science and Technology,XIDIAN University,Xian 710126,China)

机构地区：[1]中咨泰克交通工程集团有限公司,北京市100083 [2]中国公路工程咨询集团有限公司,北京市100089 [3]自动驾驶技术交通运输业研发中心,北京市100083 [4]中交集团智能交通研发中心,北京市100083 [5]西安电子科技大学空间科学与技术学院,西安市710126

出　　处：《公路》2025年第3期462-467,共6页Highway

摘　　要：针对毫米波雷达平面成像过程中,距离方位维功率谱图中弱目标被旁瓣掩盖的问题,通过分析旁瓣特性,提出了基于空间变迹的可变形卷积神经网络的旁瓣抑制方法,自适应地解决不同目标旁瓣带来的图像质量问题。仿真表明,该算法在抑制旁瓣的同时保留了弱目标和主瓣能量,并且区分了相邻的强目标。与传统旁瓣抑制方法相比,所提出算法的旁瓣抑制水平提高了约10 dB。Aiming at the problem of weak targets being masked by sidelobes in the range and azimuth power spectrum of millimeter wave radar planar imaging process,a deformable convolutional neural network sidelobe suppression method based on spatial tracking is proposed by analyzing the characteristics of sidelobes,adaptively solving the image quality problems caused by different target sidelobes.Simulation shows that the algorithm preserves the energy of weak targets and main lobes while suppressing sidelobes,and distinguishes adjacent strong targets.Compared with traditional sidelobe suppression methods,the sidelobe suppression level of the algorithm proposed in this paper has been improved by about 10 dB.

关 键 词：交通检测传感器 毫米波雷达 MIMO雷达 解速度模糊 速度补偿 

分 类 号：U491.5[交通运输工程—交通运输规划与管理] TN958[交通运输工程—道路与铁道工程]