基于拓扑结构的多元特征航迹关联算法

Algorithm of Multi-feature Track Association Based on Topology

多源传感器目标航迹关联技术是数据融合系统的核心技术之一。如果传感器定位存在偏差,势必导致基于距离统计模型及航迹相似度模型的传统航迹关联算法关联性能下降。提出了基于拓扑结构的多元特征航迹关联算法,通过提取目标特征,利用目标拓扑结构、目标航向误差及目标运动状态等多元特征参数进行航迹关联,提高了密集多目标、编队目标和漏检目标等复杂场景下的航迹关联正确率。仿真试验表明,该算法具有较高的关联正确率和鲁棒性,具备工程应用价值。

The multi-source sensor target track association is one of the key technologies of data fusion system.The performance of the traditional track association algorithm based on the distance statistics model and the track similarity model is reduced under the condition of sensor positioning deviation.The multi-feature track association algorithm based on topology is proposed,By extracting the target features and using multi-features such as the target topology,the target heading error,and the target motion state,the accuracy of track association in the complex scenes with dense multiple targets,formation targets,and missed detection targets is improved.The simulation results show that the algorithm has high correlation precision and robustness,and has value for engineering applications.