This thesis covers the design, modelling, and implementation of a disturbance rejection trajectory tracking LQR controller for a hexarotor UAV. It gives a detailed explanation of the algorithms and workflows required for implementing such a solution for any platform.
Furthermore, it describes how model uncertainty can be tested through parameter sweeps in trajectory tracking applications and investigates the performance of LQR controller schemes to deal with these uncertainties. For this, an implementation scheme is also given. Next, parameters for the UAV are estimated, and the LQR controller is tested in simulations in SIMULINK.
The results show that LQR can handle disturbances in trajectory tracking, and even that it can handle model uncertainty with varying degrees for different parameters. To finish off, concluding remarks about performance are given, and a suggestion for future work is made.