Stabilization and avoidance for unmanned vehicles via hybrid control architectures

This thesis deals with the problems of stabilization and avoidance of unsafe conditions for unmanned vehicles. We consider unmanned vehicles equipped with on-board sensors modelled as two cameras, one facing forward and one facing backward, providing measurements of the distance and misalignment to a target. For ground wheeled and marine surface unmanned vehicles we propose hybrid control laws combining the two on-board measurements and discuss stability results for the closed-loop system expressed in the on-board camera-based coordinates, using Lyapunov-based arguments. We prove robustness of the stability properties to uncertainties affecting the sensors and external perturbations acting on the vehicle. Avoidance of unsafe conditions is addressed by proposing a hybrid redesign approach, augmenting a predefined Lyapunov-based stabilizer with obstacle avoidance and input avoidance properties. Two redesign architectures are proposed, both preserving the decrease of the Lyapunov function along solutions, thus preserving asymptotic stability. Following this approach, an arbitrary number of arbitrarily placed point obstacles in the state space can be avoided, and for single-input systems an arbitrary value of the control input can be avoided along solutions. The proposed control architectures guarantee robustness of stability despite the discontinuous action required.