一种新的粒子滤波算法

被引:15
作者
王来雄
黄士坦
机构
[1] 西安微电子技术研究所
关键词
粒子滤波; 提议概率密度; 采样重要再采样; 无迹卡尔曼滤波; 跟踪;
D O I
暂无
中图分类号
TN713 [滤波技术、滤波器];
学科分类号
080902 ;
摘要
将采样重要再采样(SIR)方法与无迹卡尔曼滤波(UKF)相结合,提出一种新的粒子滤波算法.该算法具有无迹粒子滤波(UPF)粒子使用效率高和SIR粒子滤波运算速度快的优点,同时克服了UPF运算量的增长速率快于状态维数增长的缺陷.仿真结果表明,与UPF相比,本算法在几乎不影响滤波效果的前提下,大幅减少滤波所需计算量.
引用
收藏
页码:118 / 120
页数:3
相关论文
共 8 条
[1]  
Gaussian particle filter. Kotecha J H,Djuric P M. IEEE Transactions on Signal Processing . 2003
[2]  
Particle filters for posi-tioning,navigation,and tracking. Gunnarsson F,Bergman N. IEEE Trans-actions on Signal Processing . 2002
[3]  
SequentialMonte Carlo methods in practice. Doucet A,Freitas J F G de,Gordon N J. . 2001
[4]  
Novel ap-proach to nonlinear/non-Gussian Bayesian state esti-mation. Gordon N J,Salmond D J,Smith A F M. IEE Proc.Radar Signal Process . 1993
[5]  
Tracking multiple ob-jects with particle filter. Hue C,Cadre J P Le,Perez P. IEEE Transactions on Aer-ospace and Electronic Systems . 2002
[6]  
Filtering via simulation:Auxil-iary particle filters. Pitt M,Shephard N. Journal of the American Statistical Association . 1999
[7]  
Gaussian sum particle fil-tering. Kotecha J H,Djuric P M. IEEE Transactions on Signal Processing . 2003
[8]  
A tutorial onparticle filters for online nonlinear/non-Gaussian bay-sian tracking. Arulampalam S,Maskell S,Gordon N. IEEE Transactions on Signal Processing . 2002