论文部分内容阅读
对许多传感器而言,所观测到的杂波更容易集中在目标所处区域。此时,杂波不再是均匀分布,杂波的分布与真实目标所处状态相关,这与传统滤波算法中的假设不同。在此条件下,传统多目标跟踪算法的跟踪精度和实时性会受到很大影响。针对该问题,提出一种状态与杂波相关条件下的GM-CPHD滤波算法。对状态与杂波之间的相关性进行建模;根据整个监视区域的杂波分布重新计算杂波强度,并将其应用于滤波更新过程中;为降低时间复杂度,采用自适应椭球门限在算法更新步骤之前对量测集合进行预处理,使用落入门限内的量测集合进行更新步骤的运算。仿真结果表明,在状态与杂波相关条件下,本文算法较传统算法具有更好的滤波精度以及更低的时间复杂度。
For many sensors, the observed clutter is more likely to be concentrated in the area where the target is located. In this case, the clutter is no longer uniformly distributed, and the clutter distribution is related to the state of the real target, which is different from the assumption in the traditional filtering algorithm. Under these conditions, the tracking accuracy and real-time performance of the traditional multi-target tracking algorithm will be greatly affected. Aiming at this problem, a state-clutter-related GM-CPHD filtering algorithm is proposed. The correlation between state and clutter is modeled. The clutter intensity is recalculated according to the clutter distribution of the whole monitoring area and applied to the filtering update process. To reduce the time complexity, the adaptive ellipsoid threshold The measurement set is pre-processed prior to the algorithm update step and the update step is computed using a measurement set that falls within the threshold. The simulation results show that the proposed algorithm has better filtering accuracy and lower time complexity than the traditional algorithm under the condition of state and clutter.