基于毫米波雷达的机动目标跟踪算法研究  

Research on Maneuvering Target Tracking Algorithm Based on Millimeter Wave Radar

作　　者：鞠夕强 孟文[1,2] JU Xiqiang;MENG Wen(School of Mechanical Engineering,Southwest Jiaotong University,Chengdu 610031;Technology and Equipment of Rail Transit Operation and Maintenance Key Laboratory of Sichuan Province,Chengdu 610031)

机构地区：[1]西南交通大学机械工程学院,成都610031 [2]轨道交通运维技术与装备四川省重点实验室,成都610031

出　　处：《计算机与数字工程》2023年第6期1297-1301,共5页Computer & Digital Engineering

摘　　要：针对复杂交通环境下高机动目标难度大,易丢失的问题,提出一种适合机动目标的毫米波雷达跟踪流程。首先,提出一种适合毫米波雷达数据的跟踪门——基于规则的自适应跟踪门,能够根据雷达探测范围,目标属性自适应调整跟踪门。然后,根据目标高机动特性,选择匀速模型和匀加速模型为跟踪模型,利用交互式多模型算法跟踪目标。实验结果表明:交互式多模型算法较典型卡尔曼滤波跟踪精度提高18.26%;论文方法跟踪雷达目标准确度提高5.303%,虚警率下降0.813%。Aiming at the problem that tracking high maneuvering target is difficult and easy to lose in the complex traffic envi-ronment,a tracking process of millimeter wave radar suitable for maneuvering target is proposed.Firstly,a rule-based adaptive tracking gate suitable for millimeter wave radar data is proposed,which can adaptively adjust the tracking gate according to the ra-dar detection range and target attributes.Then,according to the high maneuvering characteristics of the target,the uniform velocity model and uniform acceleration model are selected as the tracking models,and the interactive multiple model algorithm is used to track the target.The experimental results show that the tracking accuracy of interacting multiple model algorithm is 18.26%higher than that of typical Kalman filter,the tracking accuracy of this method is improved by 5.303%,and the false alarm rate is reduced by 0.813%.

关 键 词：毫米波雷达 目标跟踪 交互式多模型算法 卡尔曼滤波 跟踪门 

分 类 号：U461.91[机械工程—车辆工程]