一种解耦最优分布式卡尔曼滤波算法

A Kind of Algorithm for Optimal Decoupled Distributed Kalman Filters

本文介绍一种能解耦的最优分布式卡尔曼滤波算法,它在现有的各种分布式卡尔曼滤波算法中最有特色。它应用Luenberger标准形,把一个高阶大系统分割成n个彼此没有信息交换的独立的低阶子系统,从而构成了n个独立子系统的卡尔曼德波器。这种算法最适合于应用多微处理器的分布式滤波的并行计算。由于各子系统的滤波计算是独立的,所以大大节省了滤波计算时间,提高了滤波的实时性。

An algorithm for optimal decoupled distributed Kalman filter is described in this paper. By introducing the standard Luenberger, the high - order large system is divided into several low-order independent subsystems with no in- formation exchanges. The Kalman filters are designed for these independent sub- systems. This algorithm is very suitable for the case where the distributed Kalman filters are parallel implemented on multi-processors. Since the computation of the filters for subsystems is independent,the computational time is Saved and the real- time ability of Kalman filters is increased.