连续大系统的网络结构最优滤波器

OPTIMAL NETWORK ESTIMATOR FOR LARGE SCALE CONTINUOUS-TIME SYSTEMS

本文提出一种连续大系统状态估计器的设计方法——网络结构最优滤波器。这个方法基于矩阵最小值原理,其计算结果是满足任意结构约束的最优估计。对于分散和递阶结构而言,此方法具有容错和设计灵活等特点,特別适宜于用多计算机系统来实现,且不要求信道有很宽的通频带。

The paper deals with the real-time state estimation problem of large scale systems. Based on matrix minimum principle, an optimal network estimator is proposed. The algorithms derived are proven to be optimal for large scale system state estimation subject to information flow constraints in processing net, and have flexibility in comparision with decentralized or hierarchical methods. Clearly, multicomputer systems can be available to carry out the scheme.