A vision-based nonlinear decentralized controller for unmanned vehicles

This paper presents a vision-based control strategy for decentralized stabilization of unmanned vehicle (UV) formations. The key point of the algorithm is that it only requires knowledge of the leader-follower relative distance and bearing. The approach is based on an output feedback controller that uses a high-gain observer to estimate derivatives of UV's relative positions. Both data are measured using a pan-controlled camera on-board the following robot which eliminates sensitivity to information flow among vehicles. A Lyapunov stability analysis guarantees that the closed-loop system is stable and the formation error can be made arbitrarily small. A virtual environment and a vision system are used to validate the proposed methodology.