粒子群算法在多传感器多目标跟踪的应用

多传感器多目标跟踪系统中,数据关联是其中的关键问题之一.它可以表述为多维分配问题,提出了基于粒子群优化算法的多维分配算法,它将多维分配问题中的目标代价函数极小化问题作为组合优化问题求解.通过在粒子群初始化步骤以及交叉和变异时充分考虑确认备选量测,缩小优化搜索范围,能较快找到最优解实现关联.在虚警和漏检、密集目标环境下,该算法应用于多传感器多目标融合系统仿真,结果表明所述算法在多目标数据关联中有较好的可行性和优越性.

Application of particle swarm optimization in multisensor multitarget tracking

For multisensor multitarget tracking systems,the data association problem is one of the key problems.It can be formulated as a multiple dimensional assignment problem.A multiple dimensional algorithm based on particle swarm optimization algorithm is proposed in this paper.By a combinational optimal way,it can find the minimizing solution to the objective cost function in the multiple dimensional assignment problem.The best seclusion can be searched fast in respect of reducing the search range through the validated candidate measures,which are considered in particle swarm initiation and cross rules,and mutation rules.Under false alarm,miss detection and dense targets environment,this method is used in multisensor multitarget fusion system,and the result testifies that it can solve the association problems.