Recursive Filter Algorithm for Binary-Valued Systems with Uncertain Dynamics

This paper focuses on the state estimation problem of a class of systems under binary-valued observations and uncertain dynamics. Firstly, the key idea of the extended state observer is introduced to deal with the uncertain dynamics. The idea is to regard the uncertain dynamic of the state model as a signal to be estimated, and then take it as an extended state. Secondly, based on the extended binary-valued system, an algorithm gain design method is proposed and an extended state based recursive binary-valued projection filtering algorithm is established. Furthermore, the effectiveness of the proposed algorithm is illustrated via simulations for binary-valued systems with different uncertain dynamics.