基于自适应DBSCAN的雷达目标跟踪算法

Radar target tracking algorithm based on adaptive DBSCAN

摘要:

直接对三坐标航管一次雷达点迹录取器中录取到的点迹进行航迹起始、跟踪等处理后，会形成大量虚警，运算量大.进行目标跟踪时，候选点迹集合数量庞大是造成目标跟踪过程运算量大的主要原因.文中基于动态自适应 DBSCAN 聚类算法，结合经典卡尔曼滤波跟踪算法，提出了动态自适应DBSCAN 聚类跟踪混合算法，来减少候选点迹集合数量.实验结果证明，本文提出的算法实现了无效点迹数的减少、航迹质量的提高以及运算时间的下降.通过动态自适应DBSCAN 聚类跟踪混合算法，能迅速跟踪到三坐标航管一次雷达探测到的目标并形成目标航迹，可以及时发现黑飞目标，将对正常民航飞机飞行的干扰降到最低.

Abstract:

Directly processing the track initiation and tracking of the points recorded in the radar point recorder of the threedimensional air traffic control system will generate a large number of false alarms and a large amount of calculation. When performing target tracking, the number of candidate point sets is huge is the main reason for the large amount of calculation in the target tracking process. Based on the dynamic adaptive DBSCAN clustering algorithm and the classic Kalman filter tracking algorithm, a hybrid dynamic adaptive DBSCAN clustering tracking algorithm is proposed in this paper to reduce the number of candidate point sets. Experiments have found that the number of invalid points is reduced and the track quality is improved Computing time decreases. Through the dynamic adaptive DBSCAN clustering tracking hybrid algorithm, it can quickly track the target detected by cnac radar once and form the target track, which can detect the black flight target in time and reduce the interference to the normal flight of civil aviation aircraft to the minimum.