海上低空突防群目标跟踪的IMM-Bayesian实现

由于海上低空突防编队目标存在低检测和高机动的特点,采用传统跟踪算法对编队内目标逐个跟踪存在航迹连续性差、关联混乱等问题。针对上述问题,基于对编队群整体跟踪的思想,将交互式多模型(IMM)与Bayesian算法相结合,采用IMM-Bayesian算法完成典型机动场景下(拐弯、合并、分裂)海上低空编队群目标整体的跟踪,同时利用随机矩阵作为群的扩展状态完成对群形状信息的估计。其中,对海上低空突防编队群目标运动过程中出现的分裂与合并现象,在IMM-Bayesian算法的基础上采用最近邻分类的思想对其进行有效跟踪。仿真结果表明了算法的有效性。

IMM-Bayesian Tracking Algorithm for the Sea Surface Low-Altitude Penetration Group Targets

Due to the low detection and high maneuverability of the sea surface low-altitude penetration group targets, thetraditional tracking method has the problem of poor track continuity and association confusion. Based on the idea of thewhole tracking for formation group targets, the IMM-Bayesian algorithm which combining the Interacting Multiple Model(IMM) and Bayesian algorithm was used to track the maneuvering group targets in typical scenario(turning, merging, split.ting). At the same time, the random matrix was used as the extended state of the group to complete the shape estimation.Aiming at the phenomenon of splitting and merging in the moving process of sea surface low-altitude penetration group tar.gets, the idea of nearst neighbor classification which was based on the IMM-Bayesian algorithm was adopted to track it.The simulation showed the effectiveness of the algorithm.