基于RB-GSPF算法的地形辅助导航

为了解决地形辅助导航中面临的高维、强非线性问题,提出了基于Rao-Blackwell框架的RB-GSPF算法。该算法将原系统中的线性Gauss子结构分离出来,使用经典的Kalman滤波器处理,而剩下的强非线性部分通过Gauss和粒子滤波器处理,这种结构上的分解既发挥了Kalman滤波器对于线性Gauss系统的最优性,又利用了GSPF算法结构上的优点。理论及实验分析表明:该算法与粒子滤波器相比,在降维的同时提高了定位精度,减少了粒子数目;与Rao-Blackwellised粒子滤波器(RBPF)相比,其算法结构具有更好的并行性,从而在运算量上具有优势。

Terrain aided navigation using the RB-GSPF algorithm

An RB-GSPF algorithm based on the Rao-Blackwell framework was used to solve the high dimensional,strongly nonlinear problem of terrain aided navigation.This algorithm extracts the linear Gaussian subsystem from the original system which is then solved by a Kalman filter.The other nonlinear part is handled by a Gaussian sum particle filter.This structure decomposition utilizes the optimality of the Kalman filter for a linear Gaussian system and the structural advantages of GSPF.Theoretical analyses and simulations show that the RB-GSPF algorithm has better accuracy and fewer particles than the particle filter and better parallelism and fewer computations than the RBPF algorithm.