Sočasna lokalizacija in kartiranje z robotom TurtleBot 2

Pucelj, Tilen (author), Vrabič, Rok (mentor)

V zaključnem delu je predstavljeno delovanje in uporaba algoritmov za sočasno lokalizacijo in kartiranje z algoritmi gmapping, hector mapping in cartographer. Z uporabo Turtlebot2, ki je robot s premikajočo bazo, računalnikom, kamero in deluje na osnovi robotskega operacijskega sistema (ROS), smo izdelali dvodimenzionalno karto delovnega prostora. V zaključni nalogi je predstavljena teorija robotskega operacijskega sistema, potrebna za razumevanje naloge, in delovanje naštetih algoritmov. Prikazan je postopek priprave robota in programske opreme potrebne za uporabo algoritmov v praksi. Po opravljenem kartiranju smo izmerili delavni prostor, da smo lahko preverili natančnost algoritma gmapping. Opravljena je bila tudi primerjava algoritmov v simuliranem okolju z vnaprej posnetimi podatki, ki kaže na veliko prednost algoritmov hector mapping in cartographer zaradi hitrosti kartiranja. Hkrati pokaže tudi prednost LIDAR v primerjavi z globinsko kamero za skeniranje prostora, ker nam omogoča lažji postopek kartiranja.

• Language: Slovenian
• Keywords: robotika, sočasna lokalizacija in kartiranje, dvodimenzionalni zemljevid, robotski operacijski sistem, turtleBot
• Work type: Final paper
• Typology: 2.11 - Undergraduate Thesis
• Organization: FS - Faculty of Mechanical Engineering
• Place of publishing: Ljubljana
• Publisher: [T. Pucelj]
• Year: 2020
• Number of pages: XIII, 24 f.
• PID: 20.500.12556/RUL-119964
• UDC: 007.52:004.451:912(043.2)
• COBISS.SI-ID: 32681731
• Publication date in RUL: 14.09.2020

Secondary language

• Language: English
• Title: Simultaneous localization and mapping with TurtleBot 2

Abstract

The thesis presents the operation and use of algorithms for simultaneous localization and mapping by using the algorithms gmapping, hector mapping and cartographer. Using the robot TurtleBot2, which consists of a moving base, a computer, a camera and operates on the basis of the robot operating system (ROS), we build a two-dimensional map of the workspace. In the thesis we present the theory of the robot operating system, needed to understand the given task and the operation of the listed algorithms. We show the preparation needed for the application of the algorithms in practice. After the mapping we measured the workspace and checked the accuracy of the gmapping algorithm. Lastly, a comparison of the algorithms was made in a simulated environment with pre-recorded data, which showed the advantage of using hector mapping and cartographer algorithms, because of the speed at which they map. It also displayed the advantage of using LIDAR over a depth camera for mapping, because the mapping process is less complicated.

• Keywords: robotics, simultaneous localization and mapping, two dimensional map, robot operating system, turtleBot