Izboljšanje navigacijskega sistema mobilnega robota s pomočjo referenčnih oznak

zaključna naloga Univerzitetnega študijskega programa I. stopnje Strojništvo - Razvojno raziskovalni program

Matija Zupanc (Author), Rok Vrabič (Mentor)

Abstract

Navigacija je ključni del mobilne robotike. Lahko jo razdelimo na lokalizacijo oz. iskanje položaja ter načrtovanje poti in premikanje po prostoru. Za navigacijo je razvitih že mnogo algoritmov, ki služijo vsak določenim potrebam. Kljub temu se še vedno srečujemo z izzivi. Včasih si želimo natančnejše lokalizacije, kar je včasih težko zagotoviti ali pa se pojavi težava, kot je problem ugrabljenega robota. O tem govorimo, ko robot popolnoma izgubi svojo lokacijo. Razlog se lahko skriva v neoptimiziranih ali nenatančnih algoritmih, senzorjih ali zemljevidu. Za rešitev teh težav se pojavljajo novi pristopi, ki skrbijo, da do teh težav redkeje prihaja. Eden izmed takih pristopov bo uporabljen tudi v zaključnem delu, in sicer s pomočjo programske knjižnice, ki nam omogoča zaznavo referenčnih oznak. Preko njih lahko pridemo do transformacij in jih uporabimo pri lokalizaciji robota. V delu je bilo ugotovljeno, da nam takšen sistem le malenkost izboljša navigacijo, spremembe pa se najbolj opazijo pri problemu, kadar je robot popolnoma izgubljen v prostoru.

Keywords

diplomske naloge;mobilni roboti;navigacija;ROS;Turtlebot2;AprilTag;

Data

Language:	Slovenian
Year of publishing:	2023
Typology:	2.11 - Undergraduate Thesis
Organization:	UL FS - Faculty of Mechanical Engineering
Publisher:	[M. Zupanc]
UDC:	007.52:004.93:004.45(043.2)
COBISS:	163580163
Views:	18
Downloads:	5
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Enhancing mobile robot navigation using fiducial markers
Secondary abstract:	Navigation is a crucial part of mobile robotics. It can be divided into localization or position finding, path planning and movement through space. Many algorithms have already been developed for navigation, each serving specific needs. Nevertheless, we still encounter challenges. Sometimes, we desire more precise localization, which can be difficult to provide, or an issue like a kidnapped robot arises. This refers to the situation when the robot completely loses its location. The reason could lie in unoptimized or inaccurate algorithms, sensors, or maps. New approaches are emerging to address these problems, resulting in less frequent occurrences of such issues. One of such approaches will also be used in the final thesis, using a software library that enables the detection of fiducial markers. Through these markers, we can obtain transformations and use them for robot localization. The study found that such a system only slightly improves navigation, with the most significant changes observed in situations where the robot is completely lost in space.
Secondary keywords:	thesis;mobile robots;navigation;ROS;Turtlebot2;AprilTag;
Type (COBISS):	Final paper
Study programme:	0
Embargo end date (OpenAIRE):	1970-01-01
Thesis comment:	Univ. v Ljubljani, Fak. za strojništvo
Pages:	XIV, 26 f.
ID:	19872157