基于拓扑多边形的海上船舶航迹关联方法

Correlation Method for Marine Vessel Track Based on Topological Polygon

在组网雷达存在系统误差的情况下,为解决多雷达重叠监测区域海上船舶航迹关联困难的问题,提出了基于拓扑多边形的船舶航迹关联方法。选择2部雷达的各1条航迹进行逐时刻关联,首先分别筛选单时刻下2条航迹的周围目标,通过粗匹配算法使周围目标数一致;然后分别构建拓扑多边形,再构建相似度函数计算拓扑多边形的相似性作为该时刻2个待关联航迹的相似度值;最后计算各时刻相似度的均值并根据阈值判别航迹是否关联。在不同场景的仿真实验中的航迹关联平均正确率为90.87%,在雷达存在漏检等多种场景下性能优于对比算法,为实现雷达系统误差情况下的海上船舶航迹关联提供了有效且实用的途径。

To solve the difficulty in correlating marine vessel tracks in the overlapping monitoring area of multiple radars when the radar has a system error,a method of correlation based on topological polygons is proposed in this study.One track from each radar was selected to calculate the correlation at each time point.First,the surrounding targets of the two tracks were filtered and a matching algorithm was used to keep the number of surrounding targets consistent.Second,topological polygons were constructed and a similarity function was constructed to calculate the similarity of the topological polygons as the correlation value of the two tracks to be correlated.Finally,the mean value of the correlation degree at each moment was calculated and compared with a threshold value to determine whether the tracks were correlated.In the simulation experiment,the correct mean rate of the method in this study was 90.87%,and its performance was better than that of the comparison algorithm in a variety of scenarios,such as radar missed detection.This method provides an effective and practical way to realize the track correlation of marine vessels under the condition of radar system error.