论文部分内容阅读
传统粒子概率假设密度(PHD)滤波器假定新生目标强度已知,当新生目标在整个观测区域随机出现时不再适用。为解决新生目标强度未知时的多目标跟踪问题,提出了一种基于量测信息的双门限粒子PHD(PHD-DT)滤波器。首先基于似然函数设定门限对存活目标量测进行粗提取,利用上一时刻的目标估计值构建圆形波门进行精细提取,并对门限设定方法进行分析,然后根据提取结果对目标PHD进行分解,得到存活目标和新生目标的PHD预测及更新表达式,最后给出了滤波器的实现方法并同基于量测驱动的PHD(PHD-M)滤波器和Logic+联合概率数据互联(JPDA)方法进行了仿真对比。仿真结果表明,在新生目标强度未知时,PHD-DT可有效避免Logic+JPDA在杂波背景下因航迹起始错误带来的估计误差,并较好地解决了PHD-M的目标数目过估问题,多目标估计性能更优,且杂波越强性能优势越明显。
Conventional Particle Probability Hypothesis Density (PHD) filters assume that the newborn target intensity is known and no longer apply when newborn targets appear randomly throughout the observed region. To solve the problem of multi-target tracking when the target intensity is unknown, a double-threshold particle PHD (PHD-DT) filter based on measurement information is proposed. First of all, based on the threshold of likelihood function, the survivor measurement is coarsely extracted based on the threshold, the circular wave gate is constructed by using the target estimation value of the previous moment, and the threshold setting method is analyzed. Based on the extracted result, Finally, the realization of the filter and the combination of PHD (PHD-M) filter based on measurement and Logic + Joint Probabilistic Data Interconnection (JPDA) Method of simulation comparison. The simulation results show that PHD-DT can effectively avoid the estimation error caused by the initial tracking error of Logic + JPDA in the background of clutter when the new target intensity is unknown and solve the target number of PHD-M well Estimating the problem, multi-objective estimation performance is better, and the more clutter performance advantage more obvious.