Sensor Fusion for Mobile Robot Localization using UWB and ArUco Markers

• Autor(a) principal: Silvia Dolores Nogueira Faria
• Data de Publicação: 2021
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: https://hdl.handle.net/10216/135227
• topic: Engenharia electrotécnica, electrónica e informática; Electrical engineering, Electronic engineering, Information engineering

One of the main characteristics to consider a robot truly autonomous is the fact that it is able to locate itself, in real time, in its environment, that is, to know its position and orientation. This is a challenging area that has been studied by several researchers around the world. To obtain the localization of a robot it is possible to use different methodologies. However, there are methodologies that present problems in different circumstances, as is the case of odometry that suffers from error accumulation with the distance traveled by the robot. Another problem existing in several methodologies is the uncertainty in the sensing of the robot due to noise present in the sensors. In order to obtain a more robust localization of the robot and more fault tolerant it is possible to combine several localization systems, thus combining the advantages of each one. In this work, the Pozyx system will be used, a low-cost solution that provides positioning information through Ultra-WideBand Time-of-Flight (UWB ToF) technology. It will also be used ArUco markers placed in the environment that through their identification by a camera it is also possible to obtain positioning information. These two solutions will be studied and implemented in a mobile robot, through a beacon-based localization scheme. First, an error characterization of both systems will be performed, since the measurements are not perfect, and there is always some noise in the measurements. Next, the measurements provided by the systems will be filtered and fused with the robot's odometry values by the implementation of an Extended Kalman Filter (EKF). In this way, it is possible to obtain the robot's pose, i.e position and orientation, which is compared with the pose provided by a Ground-Truth system also developed for this work with the aid of the ArUco library, thus realizing the accuracy of the developed algorithm. The developed work showed that with the use of the Pozyx system and ArUco markers it is possible to improve the robot localization, meaning that it is an adequate and effective solution for this purpose.