DESIGN OF AN IMAGE-BASED AUTONOMOUS LANDING INSTRUMENTATION AND CONTROL SYSTEM FOR UNMANNED QUADROTOR VEHICLES

• Main Author: M. Fauzan, Ihsan
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/81537
• Institution: Institut Teknologi Bandung

A quadrotor is a vehicle capable of operating autonomously and can be used for various applications. One of the challenges in quadrotor development is the landing algorithm. One commonly used landing method is image-based landing, utilizing markers, images, or patterns located at the landing site. This research aims to design a quadrotor landing system that enables the quadrotor to perform the landing process using image captures obtained from a camera. In this study, the ArUco marker, which is widely used as a landing location marker in previous research, is employed. The landing process is expected to be performed in environments with or without Global Positioning System (GPS) support. The research began with the design phase of the instrumentation system, consisting of an ArUco marker estimation system and a camera stabilization system using servo motors as actuators. The design phase continued with the development of a cascade control system with three loops to control the position. The results of this research demonstrated that the position estimation system using images was successfully tested and provided good results in terms of correlation and stability against disturbances in the quadrotor's orientation. The control system also performed well, as evidenced by successful landings both with and without GPS assistance. The quadrotor was able to land within a distance of 30 cm from the center of the ArUco marker, with a mean absolute error (MAE) of position (in cm) of (38,6422; 31,4564; 54,5046) for landings without GPS and (10,8672; 32,6929; 35,9229) for landings with GPS. For landings without GPS, improvisation was done by tuning the model obtained from flight data. Applying the tuning results during landing tests resulted in more stable landings in terms of position, although it caused fluctuations in the generated control signals and the quadrotor's position estimation. Keywords: quadrotor, instrumentation, control, landing, stabilization, image.