Exponential forgetting factor observer in discrete time

Abstract

This paper presents the development in discrete time of a solution for the observation of deterministic linear time varying systems, whose continuous time counterpart is already available in the literature under the name of exponential forgetting factor observer. This observer has a structure similar with that of the deterministic Kalman filter and relies on the same observability property as the latter. The main difference is that convergence is achieved through a different philosophy, which provides an observer with easier tuning capabilities through a single design parameter—the so-called forgetting factor.