Probabilistic Localization of a Soccer Robot
Name
Priit Kallas
Abstract
The thesis deals with the problem of localizing a mobile soccer-playing robot using Bayes filtering methods. For navigating natural environments, autonomous robots need to know where they are located even if the position of the robot is not directly observable, but rather needs to be inferred from indirect measurements of several noisy sensors. The algorithms need to account for the inherent uncertainty of such systems. Several algorithms of robot positioning including Kalman filter and particle filter are investigated, implemented and compared. The algorithms are also tested on a real robot. A working solution for practical robot localization is developed.
Graduation Thesis language
English
Graduation Thesis type
Master - Information Technology
Supervisor(s)
Konstantin Tretyakov
Defence year
2013