基于粒子群优化的Unscented粒子滤波算法

摘要

针对Unscented粒子滤波(UPF)算法中的粒子退化及重采样引起的粒子枯竭等问题,利用粒子群优化算法使粒子通过比较其当前值与最优粒子的适应度值调整自身速度,向高似然域移动,寻找最优位置,并对重采样过程进行优化,以缓解粒子的退化及枯竭问题.实验结果证明,该算法提高了UPF算法的状态估计精度.%Aiming at the problem of Unscented Particle Filter(UPF) algorithm such as particles degeneracy and particles impoverishment, by comparing particles' present values with the fitness value of objective function, it uses Particle Swarm Optimization(PSO) algorithm to make particles of UPF move towards the higher likelihood area, and finds the optimal position, and relieves the problem of particles degeneracy and impoverishment by improving re-sampling process. Experimental result proves that the state estimation precision of the improved algorithm is superior to traditional UPF algorithm.