航迹关联指派算法研究

Research on Assignment Algorithms of Track Correlation

为了解决多目标航迹关联问题,将假设检验方法和线性指派算法综合起来形成了检验优先和指派优先两类方法.针对3种典型算法在雷达探测能力受限或者目标密集场景下容易出现错误关联的问题,提出了一种改进的检验优先算法.该算法通过惩罚性地增加虚拟目标的关联代价来改善关联性能.设计最简单的二目标场景和多目标场景对该算法进行了仿真.仿真结果表明,该算法在目标密集和雷达检测概率下降时仍然能够保持较高的关联性能.

To solve the problem of multi-target track correlation, this paper forms two sorts of methods, namely test-priority and assignment-priority by integrating the hypothesis testing and linear assignment algorithms. For directing at the problem that the three typical algorithms tend to make error correlation on the occasion of the radar＇ s detection capability being limited and scene of dense targets, an improved testing-priority algorithm is proposed in this paper, where the property of correlation is boosted by increasing punitively the correlation cost of virtual targets. The simulation of this algorithm is performed by designing the simplest scene of two targets and the scene of multi-target, and the simulation results indicate that this algorithm can keep the better correlation performance while the targets are dense and the detection probability of radar is reduced.