On stable simultaneous input and state estimation for discrete-time linear systems

Abstract

This work is devoted to solving simultaneous input and state estimation (SISE) problem for discrete-time linear systems. Our aim is to develop stable SISE algorithms. By applying the minimum variance unbiased estimation technique, we derive two SISE algorithms in the presence or absence of direct feedthrough, respectively. Riccati-like equations are formulated and presented to analyze the stability conditions of the proposed algorithms. Simulation examples are provided to further illustrate the effectiveness of the proposed algorithms and support the theoretical findings. © 2011 John Wiley & Sons, Ltd.