Design and implementation of model predictive trajectory tracking controller for a quadcopter

In this study, a trajectory tracking controller was designed for a quadcopter unmanned aerial vehicle (UAV), and its performance was evaluated by performing the necessary flight tests. As the controller approach, the model predictive controller (MPC), a newly popular method in aviation, has been chosen. Firstly, nonlinear dynamic equations of quadcopter were derived by Newton-Euler approximation. Then, a control algorithm was created, and a quadcopter system was produced, together with the sensors and flight computers necessary for the algorithm to work. While creating the control algorithm, it was decided to control the angle of the quadcopter UAV with the ArduPilot flight control algorithm, which is open-source software. ArduPilot software is an algorithm consisting of cascaded PID controllers and runs on the Pixhawk flight computer. The MPC algorithm was designed for trajectory tracking by considering the derived linear equations of motion. With the simulation created in the MATLAB environment, trials were made before the real flight tests and the design parameters of the MPC were decided. The tested controller software was written in Python and run on a Raspberry Pi computer. The communication between Raspberry Pi and Pixhawk flight computers was ensured, the system was made ready for flight, and trajectory tracking tests were carried out.