基于PHD未知杂波密度多机动目标跟踪

基于随机有限集（RFS）的跳变马尔可夫系统（JMS）是多机动目标跟踪的有效方法。但现有的方法假设杂波密度是先验已知的，而实际中杂波密度是未知且可能随着环境的改变而变化。针对这一问题，提出了一种适用于线性高斯模型的未知杂波密度下多机动目标跟踪算法。该算法以未知杂波密度高斯混合概率假设密度（λ-GMPHD）滤波为基础建模杂波和真实目标，采用线性高斯JMS模型描述目标机动，推导了未知杂波密度下多机动目标跟踪的GMPHD迭代解析表达式。仿真结果表明，所提算法可实现对于杂波密度以及目标数和目标状态的准确估计。

Multiple maneuvering targets tracking with unknown clutter density using PHD

The jump Markov system (JMS) based on the random finite set (RFS) is an effective approach for multiple maneuvering targets tracking. However, these approaches assume that the clutter density is known and priori. This is unrealistic for real applications, as it is often previously unknown and its value may be time-varying as the environment changes. To solve this problem, this paper proposes a novel algorithm for multiple maneuvering targets tracking with the linear Gaussian models in the case of unknown clutter density. The proposed method models the clutters and actual targets based on the Gaussian mixture probability hypothesis density filter with unknown clutter rate (λ-GMPHD), which removes the need of the prior clutter density, describes the maneuvering process by the linear Gaussian JMS and derives a closed-form solution to the GMPHD recursion for multiple maneuvering targets tracking under unknown clutter density. The simulation results indicate that the proposed algorithm can accurately estimate the target number and corresponding multi-target states as well as the clutter density.