基于模式空间的群目标合并与分离方法研究

A Study of Merging and Splitting Methods for Group Targets Based on Pattern Space

针对飞行编队形成的群目标航迹关联的特殊性,提出了群目标合并与分离涉及的4要素:位置、方向、机动曲率和航迹历史。在LHong构建的模式空间(即将量测数据序列由量测空间经小波变换映射到模式空间)的基础上,进一步提出了在模式空间利用其位置、指向、机动特征和航迹历史确定群目标合并与分离的新方法,即确定群目标成员的合法性方法。在模式空间经过Kalman滤波后再将模式空间变换到量测空间,以输出正确的量测。仿真结果证明了该算法的正确性与有效性。

Group-target track association is a special problem as compared with single target track association,because there are many echoes occur in every scanning period.Four essential elements involved in the merging and splitting of group targets,including location,pointing,maneuver curvature and track history,are proposed.On the basis of the pattern space developed by L.Hong,a new method of merging and splitting of group targets in pattern space is proposed using the four elements.By converting the pattern space to measurement space after Kalman filtering,correct measurements are output.Simulations show the correctness and effectiveness of the method for merging and splitting of group targets.