Akademik Veri Yönetim Sistemi
Tezin Türü: Yüksek Lisans
Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, Havacılık ve Uzay Mühendisliği Bölümü, Türkiye
Tezin Onay Tarihi: 2014
Öğrenci: EMRE CAN SUİÇMEZ
Danışman: ALİ TÜRKER KUTAY
Özet:In this thesis, trajectory tracking of a quadrotor UAV is obtained by controlling attitude and position of the quadrotor simultaneously. Two independent control methods are used to track desired trajectories accurately. One of these methods is a nonlinear control approach called as "backstepping control". The other method is a more unique optimal control approach called as "Linear Quadratic Tracking(LQT)". In addition, fixed-gain LQR controller which is widely used in literature is also used for comparison analysis. First, nonlinear dynamic model of quadrotor is obtained by using Newton’s equations of motion. Then, backstepping controller is obtained in three steps and simulation model of the backstepping controller is formed. On the other hand, time-varying optimal control gains of LQT controller are found offline by solving matrix difference Riccati equation(DRE) backwards in time. Then, LQT controller is modeled by using time-varying optimal control gains as a state feedback controller. Several trajectories to be followed are generated in MATLAB and sent into the simulation models as inputs. Finally, backstepping, LQT and LQR controllers are simulated in MATLAB/Simulink environment, for inital validation. Several trajectories are tried to be followed by each controller and simulation results of controllers are compared to each other. It is observed that, LQT controller could track relatively complex trajectories more accurately and efficiently compared to backstepping and LQR controllers. Other advantageous and disadvantageous characteristics of each control method are also analyzed in details. In this thesis, "AscTech Hummingbird" quadrotor manufactured by Ascending Technologies is used. "AscTech Hummingbird" quadrotor gives opportunity to test high level control algorithms generated in MATLAB/Simulink environment. Therefore, complete validation of controllers obtained in this thesis could be performed by real time experiments in future.