【作者】 张钰； 林雨田； 高利； 赵亚男； 吴绍斌；

【Author】 Zhang Yu;Lin Yutian;Gao Li;Zhao Yanan;Wu Shaobin;Beijing Institute of Technology;

【机构】 北京理工大学；

【摘要】 毫米波雷达对于远距离静态物体检测能力较弱,在识别前方静态物体时目标丢失和信息缺失现象时有发生,驾驶辅助系统所搭载的毫米波雷达对静态物体的误检率和漏检率也相应上升,严重降低了驾驶辅助系统的安全性和可接受性。针对毫米波雷达对于静态物体辨识能力较弱,本文使用了机器学习算法来实现静态物体检测和跟踪。首先,远处静止障碍物,车辆以一定速度接近,采集并处理分析不同工况下毫米波雷达的数据,结果表明毫米波雷达在静态物体距离超过100m并且车速超过19m/s时的检测能力较差。然后,选取雷达检测目标的相对速度和相对距离作为观测量,使用高斯隐马尔可夫模型学习雷达的标签数据,获取了相对速度、相对距离和目标状态之间的非线性关系。基于雷达信号,通过高斯模型聚类可以对目标标签进行预测,再通过前后向反馈算法实现目标跟踪。结果表明,本文使用的模型能够有效地改善雷达对于距140m远、速度达到30m/s物体的目标预测和跟踪精度。更多还原

【Abstract】 Due to the poor ability of millimeter wave radar in recognizing distant static objects,target loss and incomplete information will occur when it recognizes the static target in front,thus increasing the false alarm rate and missing alarm rate of the radar-dependent driving assistant system,which will reduce the driving safety and the acceptability of the assistant system.Aiming at the radar’s poor ability to recognize static targets,this paper uses machine learning algorithm to recognize and track targets.Firstly,test vehicle approaches distant static target with constant velocity,and the radar signals are collected and processed in different conditions.The results show that the radar has a poor recognition effect when the distance is more than 100 meters and the vehicle speed is more than 19 m/s.After that,taking the relative distance and relative velocity of the radar target as observation quantity,Gaussian Hidden Markov Model(GHMM) is used to learn the labeled data from radar,and the nonlinear relationship among the relative distance,relative velocity and target states are obtained Moreover,through the clustering of Gaussian Model(GM),the target’s label can be predicted according to the radar signal,and the forward-backward algorithm can be used to track the target.The results show that the model can effectively improve the accuracy of target prediction and tracking effect when the distance is around 140 meters and the speed is around 30 m/s.更多还原