Modeling, identification and control of a quad-rotor drone using low-resolution sensing

Modeling;Identification;Control;quadrotor;AR.Drone

Sun, Yue ; Dullerud ; Geir E.

关键词: Modeling;    Identification;    Control;    quadrotor;    AR.Drone;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/34205/SUN_YUE.pdf?sequence=1&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

This thesis focuses on the modeling and identification, control and filter design, simulation and animation, and experiments of an electrical-motor drive model-scale quadrotor --- the AR.Drone. Equations of Motion of drone’s model were derived from Kinemics and Dynamics of common quadrotors. The identification was conducted thoroughly including its low-resolution on-board sensors, such as rate gyro and altimeter. Control targets are composed of two stages --- local references following and global position tracking. PID algorithm is used by both controllers with various filters designs, such as low/high pass filter, Complementary Filter and Kalman Filter. Simulation is also divided to two stages with two different simulators ---- MATLAB and C++. The first stage MATLAB simulation is intended to only test the controllers with no disturbances or noises. The second stage high fidelity C++ simulation contains everything including animation. Experiments results are presented and correlated to simulation to evaluate the identification and modeling. This thesis also includes modeling and identification of a low-resolution camera sensor --- Kinect. The model is included in global position tracking simulation. Some experiments videos and animation videos are available at http://www.youtube.com/user/sunyue89/videos.The author hopes this thesis is helpful to researchers and amateurs who would like to develop the AR.Drone or any other small scale quadrotors using low-resolution sensing for autonomous control.