空间密集群目标关联与跟踪算法研究

Study on Correlation and Tracking Algorithm of Space Dense Group Targets

为提高对空间密集群目标跟踪的精度,提出了基于全局最近邻的群目标关联与跟踪算法。通过对空间密集目标进行群分割,将群跟踪问题转化为多目标跟踪问题,考虑到空间目标具有运动速度快的特性,基于“全局最优”原则选取距离最近的群目标和量测进行优先关联与更新,避免关联冲突和减少关联错误,可有效解决关联准确性与跟踪实时性之间的矛盾,同时提出航迹预测与轨迹预报相结合的方法,来解决跟踪过程中的航迹断续与融合问题。仿真实验结果验证了所提算法的有效性。

In order to improve the accuracy of target tracking in spatially dense clusters,an algorithm based on the global nearest neighbor principle is proposed.The group tracking is transformed into multi-target tracking by group segmentations.Considering the fast motion speed of the space targets,the nearest group targets and measurements are selected to prioritize association and track upgrading based on the principle of“global optimal”.In this way,correlation conflicts can be avoided and correlation errors can be reduced,and the contradiction between correlation accuracy and real-time tracking performance can be effectively solved.At the same time,a method combining track path prediction and trajectory prediction is proposed to solve the problem of track path discontinuity and fusion in the tracking process.At last,the simulation results verify the effectiveness of the algorithm.