Visual Motion Observer-based Pose Control with Input Saturation

Abstract

This paper investigates the visual feedback control by the visual motion observer with input saturation. Firstly, using standard body-attached coordinate frames (the world frame, camera frame and object frame), we present the visual motion error system which consists of the estimation error system and the pose control error system. Next, we propose the control law with the input saturation. After that, stability analysis of the closed-loop system is discussed in the sense of Lyapunov. Although the proposed control law cannot be designed based on the passivity of the visual motion error system explicitly, the skew-symmetric property of the visual motion error system plays an important role in the stability analysis. Finally, experimental results are shown in order to confirm the proposed method by using AR.Drone as a small unmanned aerial vehicle.