并行局部概率假设密度(PHD)粒子滤波

针对概率假设密度滤波权值更新效率低下的问题, 提出一种并行局部概率假设密度粒子滤波?通过聚类将粒子按目标估计个数进行分类并添加标签,通过一步预测和跟踪门限将观测区域划分为目标存在区域和不包含目标的杂波区域,修正目标所在的局部区域杂波强度公式,独立并行地计算每个目标所在区域的局部概率假设密度?仿真结果表明,并行的局部概率假设密度粒子滤波时效性更高,误差更低?

Parallel local PHD particle filter

To improve the effectiveness of probability hypothesis density filter when to update weights,a parallel local probability hypothesis density particle filter is proposed.Firstly,all the particles are classified into several types equivalent to the number of targets and labelled through clustering algorithm.Then observation region is partitioned into target parts and clutter parts. Last,the clutter intensity of region in targets''gate is modified and local probability hypothesis density in every targets''gate is computed parallely and simultaneously.Simulation results show that,compared with probability hypothesis density particle filter,parallel local probability hypothesis density particle filter has higher timeliness and lower error.