Asservissement et Navigation Autonome d'un drone en environnement incertain par flot optique
Bruno Hérissé · HAL (Le Centre pour la Communication Scientifique Directe) · 2010
This thesis deals with navigation of a VTOL Unmanned Aerial Vehicle in unknown or uncertain environment. The use of optical flow is bio-inspired. It provides information on velocity of the vehicle and proximity of obstacles. Two contributions are presented in this work. The first one addresses automatic landing on a static or mobile platform. The maneuver is split into two tasks : stabilization of the linear velocity above the target and vertical landing. The approach shows that regulation of the divergent optical flow around a constant set point enables a smooth landing without collision despite uncertainties on the vehicle and the platform dynamics. The second contribution concerns terrain following with obstacle avoidance. The general approach considered enables us to address different applications such as corner avoidance, steep terrain following, corridor following, etc. Stability analysis assesses robustness and limits of the controller in the presence of various uncertainties such as uncertainties on the geometry of the environment. All the algorithms are simulated and experimented on a quadrotor UAV built at the CEA LIST.