基于矩阵递推求逆的快速联邦Kalman滤波

Rapid federal Kalman filtering based on the inverse matrix recursive

在SINS/GPS组合导航系统中大多采用联邦Kalman滤波算法,但联邦Kalman滤波算法存在大量矩阵相乘运算和矩阵求逆运算,从而影响了算法的运行速度。为了解决此问题,在联邦Kalman滤波算法子滤波器中引用矩阵外积方法对矩阵相乘进行处理,并提出了基于矩阵递推求逆的快速联邦Kalman滤波算法应用于主滤波器中,从而大大提高了算法的运行速度。经理论分析和实验结果表明,矩阵处理后的新算法计算运行所用时间是原算法的1/4,精度也有所提高。

In the SINS/GPS navigation system for most use federal Kalman filtering algorithm,but the existence of matrix multiplication operation and inverse matrix,which influence the running speed of the algorithm.In order to solve this problem,in federal Kalman filtering is referenced in the way filter matrix method of matrix multiplication deposition,and puts forward the solution of recursion inverse matrix based on the rapid Kalman filtering algorithm is applied to the Lord in the filter algorithm,thus greatly improve the speed.Through theoretical analysis and experimental results show that the new algorithm is processed matrix computation time is running one quarter of the original algorithm,and also improved accuracy.