Modelagem e controle de um quadricóptero parrot rolling spider

Abstract

Autonomous Unmanned Aerial Vehicles (UAVs) have wide applications, in the civil or military field, which contributes to reducing costs and risks. This work presents the application of modeling and control techniques of a Rolling Spider model quadcopter, manufactured by the French company Parrot. Two control approaches were used - the Linear Quadratic Regulatory Controller (LQR) and the Proportional-Integral-Derivative (PID) Controller - for stabilization and tracking of the vehicle. The mathematical modeling of the kinematics and dynamics, for the rotation and translation movements, of the quadcopter was carried out using the Newton-Euler formalism. The non-linear model was obtained after defining the relevant states for a faithful simulation of the vehicle dynamics. Next, a numerical linearization of the plant was carried out, in relation to an equilibrium point for hovering flight. For the simulations of the models used, a simulator was implemented using the Matlab/Simulink software, and the responses were analyzed in relation to the attitude commands, which refer to the Euler angles used to control the inclination, rotation, and orientation, in addition to the command vehicle altitude. A comparative analysis of altitude control is made using PID and LQR controllers. For a given reference trajectory, the LQR controller presented a stabilization time of 2 s, and a steady state error of 0.005, while the PID controller presented, respectively, 5 s stabilization time and a steady state error of 0.63. The results validate that the LQR controller is more efficient for altitude control in Parrot Rolling Spider quadcopters.