杂波环境下多机动目标跟踪的一种新粒子滤波器算法研究

多机动目标跟踪系统具有本质非线性和非高斯特性, 而且状态的测量通常是来源未确定的观测量, 在杂波环境下尤其如此.因此, 杂波环境下的多机动目标跟踪问题需要同时解决多个目标的机动性以及测量和目标之间的关联问题.提出了一种结合粒子滤波器和吉布斯采样器的多机动目标跟踪算法, 可以很好地解决在杂波环境下的多机动目标跟踪问题.在仿真研究中, 与另一多机动目标粒子滤波器算法BMMPDAF的比较证明了所提出的算法的优越性.

A New Particle Filter for Multiple Maneuvering Target Tracking in Clutter

Multiple maneuvering target tracking systems are intrinsically nonlinear and non-Gaussian,and the measurements are usually unresolved, especially in cluttered environment. So tracking multitarget with maneuverability in clutter has to tackle the maneuvering of multitarget as well as the association between measurements and targets. So a new multitarget tracking algorithm is proposed, which combines the particle filter and the Gibbs sampler, and can track maneuvering multitarget in cluttered environment very well. In a simulation comparison with another multitarget tracking algorithm, BMMPDAF, the priority of the proposed algorithm is presented.