一种改进的基于蚁群优化的粒子滤波算法

摘要

This paper proposed an improved particle algorithm based on ACO(ant colony optimization)to solve the problem of particle deficiency and loss of particle diversity in traditional particle filter algorithm.The improved algorithm optimized the resampling process of particle fiber algorithm by ACO,the small weight particles moved to the location with better weight particles by transition probability after the weights updating,which prevented the smaller weights particles disappear after several iterations,at the same time,the algorithm set a transition threshold to refrain transfer between better weights particles.They solved the problem of particle deficiency and loss of particle diversity at the same time.Experimental results show that this algorithm has higher states estimation accuracy and better robustness.%针对传统粒子滤波算法中粒子匮乏以及粒子多样性丧失的问题,提出了一种基于蚁群优化的改进粒子滤波算法.该算法利用蚁群算法优化粒子滤波的重采样过程,使粒子在更新权值后,利用转移概率向权值较优粒子的位置移动,以防止权值较小的粒子在多次迭代后退化消失；同时,设置转移阈值,以抑制权值较优粒子间的转移,从而同时解决了粒子匮乏以及粒子多样性丧失的问题.实验结果表明,该算法具有较高的预估精度和较好的鲁棒性.