机动目标跟踪系统通常是非线性而且不完全观测的 ,所以问题的关键在于每一时刻的目标机动性都是高度不确定的。提出了一种新的平滑粒子滤波算法 ,该算法在粒子滤波器中加入了对系统模型的概率分布密度的平滑处理 ,从而很好的解决了目标的机动性估计问题。在仿真研究中 ,与辅助粒子滤波器的比较验证了本文算法处理非线性跟踪问题的优越性 The maneuvering target tracking system is usually nonlinear and not completely observed, so the crux of the problem is that the target maneuver of each moment is highly uncertain. A new smoothing particle filter algorithm is proposed. The proposed algorithm smoothes the probability distribution density of the system model in the particle filter, which solves the maneuvering problem of the target well. In the simulation study, the comparison with the assistant particle filter verifies the superiority of our algorithm in dealing with the nonlinear tracking problem