Descent Control System of a Quadcopter Based on Fiducial Markers

• Grant number: 24/18500-4
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: January 01, 2025
• End date: December 31, 2025
• Field of knowledge: Engineering - Electrical Engineering - Industrial Electronics, Electronic Systems and Controls
• Principal Investigator: Roberto Santos Inoue
• Grantee: Lucas Bosso de Mello
• Host Institution: Centro de Ciências Exatas e de Tecnologia (CCET). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil

Abstract

It is known that the Unmanned Aerial Vehicles (UAVs) perform a wide variety of tasks, which often requires the aircraft to land during operation. This can be observed in material delivery operations, for example. In this scenario, it is important that the landing is carried out with high precision, ensuring the safety of the environment and the drone. In this research project, a landing pack will be developed to control the positioning of the quadcopter during landing, aiming to achieve highly precise landings on the platform point desired by the user. For this purpose, fiducial markers will be used for altitude estimation and an attached camera for their recognition. Furthermore, the drone has a GPS and odometry system, which will provide information for a fusion through a Kalman filter with the altitude measurement from the fiducial marker. The fusion will provide a more accurate altitude estimate for the drone, allowing the descent controller to perform a smoother landing.The algorithm will be developed using the ArUco fiducial marker, which offers high detection by cameras and low computational cost. To develop the landing package, the bebop_autonomy driver for the Robot Operating System (ROS) framework will be used with a quadcopter, the Parrot Bebop 2. The Kalman filter will be used to provide a better estimate of the drone's position based on odometry and GPS, and with this estimated positioning , a PID controller will be applied to manage the descent and center the aircraft on the platform. Finally, for final validation, the algorithm will be applied directly to the drone, aiming to achieve precise landings on the developed platform. This precision will be statistically analyzed through the quadcopter's actual final position and the predicted position.