杂波环境下基于IMM-PSNF的机动目标跟踪算法

Maneuvering Target Tracking in Clutter Environment Based on IMM-PSNF

鉴于现有算法对杂波环境下的机动目标跟踪,在杂波密度较大、机动性强的情况下滤波器容易发散且算法复杂度较大的问题,提出了交互式概率最强邻算法(IMM-PSNF)。该算法以传统的机动目标跟踪算法IMM为基础,利用算法复杂度较小的概率最强邻算法(PSNF)进行滤波更新,并借助两级模型概率加权波门技术增强滤波器的稳健性。理论分析和仿真结果都表明,通过最强邻思想对量测的有效选择,IMM-PSNF无论是跟踪精度还是航迹丢失率都要优于全邻滤波器IMM-PDAF。

For existing algorithms for maneuvering target tracking in clutter environment,it is easy to divergence because of the clutter density and the target mobile and they are usually more complex.So this paper presents an interactive probabilistic strongest neighbor algorithm(IMM-PSNF).The algorithm based on the traditional IMM algorithm for maneuvering target tracking and used a probability of strongest neighbor algorithm filtering(PSNF)which is small complexity to update.Moreover,the algorithm uses the two-stage model probability weighted gate to enhance robustness.Theoretical analysis and simulation results show that,by selecting the most effective measure,IMMPSNF both tracking accuracy and percentage of lost tracks is better than IMM-PDAF.