Aplicação de controle ótimo na estabilização de veículo aéreo não tripulado quadricóptero limitado a um grau de liberdade

Abstract

There is a vast literature on the use of classical control (PID) in unmanned aerial vehicles (UAVs). This work, in turn, seeks to explore the application of modern control, specifically optimal control, in this type of vehicle. Using a low cost prototype created for this project, tests about the stabilization of one degree of freedom will be conducted with each control type. Along this work, it will be presented the mathematical model identification process, the control application, and experiments done with the intent of comparing the conventional PID control with optimal control. Through these tests, benefits of the modern control have been verified, such as the reduction of the cost function, lower settling time, lower overshoot, and better reaction to mechanical disturbs. The results comproved that, despite the complexity of the identification process for this kind of vehicle, optimal control can be a viable option for performance improvement for quadcopter UAVs, resulting in benefits over the traditional PID control.