融合RTS平滑的迭代无迹卡尔曼粒子滤波方法

摘要

This paper proposed a new proposal distribution function for the particle filtering,the Iterated Unscented Kalman Filter and the RTS smoothing are fused to generate the proposal distribution function so as to reduce the phenomenon of particle′s shortage.Compared with the particle filtering with only the proposal distribution function of Unscented Kalman Filter,this method can get more accurate estimation results and the system has more stability. Finally,the simulation results proved the effectiveness of the proposed method.%对粒子滤波算法中建议分布函数的设计提出了一种新的方法，即将迭代无迹卡尔曼滤波（Iterated Unscented Kalman Filtering，IUKF）与Rauch-Tung-Striebel（RTS）平滑算法融合，产生新的建议分布函数，以减小粒子滤波的粒子数匮乏现象，与单独使用无迹卡尔曼滤波产生建议分布函数的粒子滤波方法（Unscented Kalman Particle Filtering，UPF）相比，状态的估计结果更加准确，系统具有更好的稳定性。最后通过仿真研究验证了该方法的有效性。