一种基于改进蚁群算法的多目标跟踪数据关联方法

A DATA ASSOCIATION METHOD FOR MULTI-TARGET TRACKING BASED ON IACA

针对目标跟踪数据关联问题给出一种基于改进蚁群算法的数据关联方法。首先,根据多目标数据关联问题的特点,将该问题转化为组合优化问题;其次,将精英策略和排序策略引入传统蚁群算法,得到改进的蚁群算法,利用蚁群算法解决组合优化问题的优势,将改进的蚁群算法应用于多目标跟踪数据关联中,建立数据关联模型并给出基于改进蚁群算法的数据关联方法;最后,通过仿真实验验证了所获的基于改进蚁群算法的多目标数据关联方法的有效性和优势。

For the data association problem of multi-target tracking,a data association method based on the improved ant colony algorithm（IACA） is given in this paper.Firstly,according to the characteristics of multi-target data association issue,we convert this issue to combinatorial optimisation problem.Secondly,by introducing elite strategy and sorting strategy into the traditional ant colony algorithm（ACA）,we obtain an improved ant colony algorithm.By making use of the advantage of ACA in solving combinatorial optimisation problems,we apply the IACA to multi-target tracking data association,construct the data association model,and give the IACA-based data association method.Finally,through simulation experiment we verify the effectiveness and advantage of the obtained IACA-based multi-targets data association method.