Control strategies of a tilt-rotor UAV for load transportation

• Ano de defesa: 2014
• Autor(a) principal: Marcelino Mendes de Almeida Neto
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais
• Palavras-chave em Português: Engenharia elétrica; Aeronave não tripulada; Controle H; Linearização por realimentação de Saída; Seguimento de Trajetórias; Filtro de Kalman; Tilt-rotor; Transporte de carga; Veículos aéreos não iripulados
• Link de acesso: https://hdl.handle.net/1843/BUOS-9Q6GFQ

This dissertation presents control strategies to solve the problem of suspended load transportation by a Tilt-rotor Unmanned Air Vehicle (UAV) passing through a desired trajectory. For the present study, it is important for the aircraft to maintain itself and the load stable even in the presence of external disturbances, parametric uncertainties and measurement errors. In general, a precise dynamic model of a system is needed in order to design advanced control strategies to it. Therefore, a rigorous dynamic model is derived for the Tilt-rotor UAV with suspended load using Euler-Lagrange formulation. After obtaining the model, it is then possible to design control laws that satisfy the desired specifications. Consequently, linear and nonlinear control laws are designed. In order to design linear control laws, the system is linearized around its operation point. Two linear control laws are designed: one using D-stability control design and the second using simultaneous D-stability and minimization of the H1 norm. As for the nonlinear control design, a three-level cascade strategy is proposed. Each level of the cascade system executes a control law through the method of input-output feedback linearization. Each one of these levels controls a different group of the system's state variables until the aircraft becomes fully stable. Two path tracking controllers are specified for this strategy. The first considers the load only as a disturbance and does not actuate to avoid its swinging. The second controller, on the other hand, seeks to find a compromise between path tracking and reducing the load's swing. At last, as proof of concept, the nonlinear strategy is modified so that the aircraft is able to stabilize an inverted pendulum. For all the described control laws, it is considered that the physical measurements of the aircraft are precisely known in all time instants. However, every physical measure is subject to errors and uncertainties and one cannot always obtain a high sampling frequency when measuring process variables. Therefore, part of this work is dedicated to the study of uncertainties when measuring the position of the aircraft in a situation where the controller has a higher sampling frequency than the GPS. In face of this problem, the aircraft's position must be estimated while no new measurements are available taking also into the consideration the existence of disturbance inputs on the system. This whole problem is solved by using the Kalman Filter with Unknown Inputs.