多传感器集中式观测融合Kalman滤波器快速算法

对多传感器线性离散时变随机系统,虽然基于Riccati方程的集中式观测融合Kalman滤波器算法可给出全局最优状态估计,但其缺点是要求计算高维逆矩阵,计算负担大。为了克服这个缺点,应用信息滤波原理,基于改进的Riccati方程,或逆预报误差方差阵方程,或逆滤波误差方差阵方程,提出了相应的全局最优集中式观测融合Kalman滤波器的三种快速算法,可避免高维逆矩阵,可明显减小计算负担,便于实时应用,一个数值仿真例子说明了它们的有效性。

Fast Algorithms for Multisensor Centralized Measurement Fusion Kalman Filter

For the linear discrete time-varying stochastic systems with multisensor, although, the centralized measurement fusion Kalman filter algorithm based on the Riccati equation can give the globally optimal state estimation, but its drawback is to require the computation of high dimensional inverse matrix, which yields a large computational burden. In order overcome this drawback, using the information filtering principle, based on the modified Riccati equation, or inverse prediction error variance mattix equation, or inverse filtering error variance matrix equation, the corresponding three fast algorithms for globally optimal centralized measurement fusion Kalman filter, are presented, which avoid the high dimensional inverse matrix, and can obviously reduce the computational burden, and are suitable for real time applications. A numerical simulation example shows their effectiveness.