摘要
针对传统基于随机集的多目标跟踪方法高斯混合概率假设密度滤波器在强杂波环境下,会出现目标数目过估计及目标状态估计误差急剧增大的问题,提出了一种适用于强杂波密度环境下的多目标跟踪算法。该算法在概率假设密度平滑器基础上,在预测步通过量测与预测值之间的残差值确定椭圆门限值,从而获得与目标真实状态相近的量测值,降低了滤波器的时间复杂度;同时,在使用有效量测对高斯项进行更新的过程中,利用各高斯项的椭球门体积自适应调整更新公式中杂波强度参数,提高了多目标跟踪精度。仿真结果表明:该方法在强杂波密度环境下可有效降低计算时间并提高多目标跟踪精度。
Multi-target tracking in strong clutter often faces the problem of dense traces.The traditional multi-target tracking algorithm based on random finite set Gaussian mixture probability hypothesis density filter will overestimate the number of targets and sharply increase target state estimation errors.To solve the problems,a multi-target tracking algorithm under strong clutter density environment is proposed.Based on the probability hypothesis density smoother,the algorithm determines the ellipse threshold through the residual value between the measurement and the predicted value in the prediction step,so as to obtain a measurement value close to the true state of the target and reduce the time complexity of the filter.In the process of updating the Gaussian term with the effective measurement,the clutter intensity parameter in the update formula is adaptively adjusted by the ellipsoidal gate volume of each Gaussian term to improve the accuracy of multi-target tracking.Simulation results show that this method can effectively reduce the calculation time and improve the accuracy of multi-target tracking under strong clutter density environment.
作者
刘政玮
陈映
鲁耀兵
LIU Zhengwei;CHEN Ying;LU Yaobing(Beijing Institute of Radio Measurement,Beijing 100854,China)
出处
《现代雷达》
CSCD
北大核心
2022年第2期16-22,共7页
Modern Radar
关键词
多目标跟踪
概率假设密度滤波器
强杂波密度
平滑器
multi-target tracking
probability hypothesis density filter
strong clutter density
smoothing filter
