Ver Post

Apresentação publica de proposta de Tese - CAT

2 Fevereiro 2016, 09:41 - Maria João Silva Carvalho

Candidato: Pedro Daniel Graça Casau

Tìtulo: Hybrid Control for Unmanned Air Vehicles

Orientação: Carlos Jorge Ferreira Silvestre

                   Ricardo Gabriel Sanfelice


3 de Fevereiro de 2016 às 18h, Sala de Reuniões do DEEC, 5º piso Torre Norte


Abstract: This thesis explores the applications of hybrid controller synthesis to Unmanned Air Vehicles (UAVs). Firstly, we develop a hybrid control strategy that provides autonomous transition between hovered and leveled flights to a model-scale fixed-wing aircraft. The aircraft closed-loop dynamics are described by means of a Hybrid Automaton with the Hover, Transition, Level and Recovery operating modes, each of which corresponding to a different region of the flight envelope. Linear Parameter Varying control techniques are employed in Hover and Level, providing robust local stabilization, and a nonlinear locally Input-to-State Stable (ISS) controller provides practical reference tracking to the Transition operating mode. These controllers, together with an appropriate choice of reference maneuvers, ensure that a transition from hovered flight to level flight, or vice-versa, is achieved. Whenever the aircraft state reaches unexpected values, the Recovery controller is triggered in order to drive the aircraft towards stable hovered flight, providing a chance to retry the transition maneuver. The controller’s performance and robustness is assessed within a realistic simulation environment in the presence of sensor noise and wind.
Secondly, we address the problem of designing a control law based on sensor measurements that provides global asymptotic stabilization to a reference trajectory defined on the special Euclidean group of order three. The proposed control law is a function of the angular velocity, vector measurements characterizing the position of some given landmarks and of their rate of change. We provide sufficient conditions for the existence of synergistic potential functions that are used in the generation of a suitable hybrid control law. We also provide sufficient conditions on the geometry of the landmarks to solve the given problem. The proposed solution is simulated and compared with a continuous feedback control law.
We tackle the problem of trajectory tracking for a particular class of underactuated vehicles with full torque actuation and a single force direction (thrust), which is fixed relative to a body attached frame. Additionally, we consider that thrust reversal is not available. Under some given assumptions, the control law that we propose is able to track a smooth reference position trajectory while minimizing the angular distance to a desired orientation. This objective is achieved robustly, with respect to bounded state disturbances, and globally, in the sense that it is achieved regardless of the initial state of the vehicle. The proposed controller is tested in an experimental setup, using a small scale quadrotor vehicle and a motion capture system. Finally, we address the problem of globally uniformly exponentially stabilizing a linear system to the origin by output feedback while avoiding a prescribed set of input values. We provide sufficient conditions for avoiding a set of input values within a polyhedral and we show that, for the particular case of a polytope, the problem can be solved using two controllers and appropriate switching between them. The proposed hybrid controller can be adapted to solve the problem of global asymptotic stabilization of a quadrotor vehicle without the limitations of saturated feedback and the lack of robustness to exogenous perturbations.