Formation flight of UAV\'s with obstacle avoidance

• Ano de defesa: 2022
• Autor(a) principal: Rosa, Victor Stafy Megda
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: autonomous flight; control; controle; desvio de obstáculos; dinâmica de voo; flight dynamics; formation flight; obstacle avoidance; UAV; VANT; voo autônomo; voo em formação
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/18/18161/tde-13102022-092006/

This thesis presents a control methodology to address the problem of formation flight of autonomous aircraft with a control system capable of avoiding obstacles. A literature review was performed to explore some of the essential fields needed to implement a control system that can solve this problem. The main configurations of formations of autonomous vehicles were studied, as well as the principal control techniques available in the literature, and applied to formation flight. The sensors for UAV navigation and sensors for obstacle identification were also studied. Eventually, some results obtained from the literature were presented, which show benefits coming from formation flight, like induced drag reduction for the aircraft set. Also, some formation geometries like V formation, Echelon, Chevron, and inverted V were studied. A quadrotor model was developed, which is used as the standard model for aircraft formation. The theory used to perform model predictive control is also presented, as well as the technique used to track the desired planned trajectory. Moreover, it is described how the sensors, disturbance, and noise were simulated, as well as the techniques and modifications implemented in the control methodology to perform obstacle avoidance, make use of soft constraints, and improve algorithm runtime. Eventually, it is presented the simulation results using the quadrotor model previously developed. In these simulations, the aircraft perform formation flight in an obstacle-laden environment. The simulations included disturbance and noise effects as well as other real-world issues. The main contribution of this work is the establishment of a new controller design methodology using the MPC technique applied to the formation flight with obstacle avoidance problem. Such that the nonlinear aircraft model as well as the uncertainty, noise, disturbance, and other problems, which occur during an aircraft\'s mission are taken into account.