[Journal papers] [Conference papers] [PhD Theses] [MSc Theses]


PhD thesis
Autonomous Navigation of Mobile Robots Based on Ultrasound Range Sesnsors
Faculty of Electrical Engineering and Computing, University of Zagreb, Croatia, 2009 ((in croatian), Supervisor: Professor Ivan PetroviŠ)
Mobile robotics is in the last few decades a very dynamic area of development.
Mobile robots become more and more important in industry, science and everyday
human life. Research in this area can be divided in two levels: first level is about
low level mobile robot control, which includes mobile robot navigation, localiza-
tion, map building and planning of a safe path through a working environment
with various stationary and moving obstacles, second level is about improving the
mobile robot autonomy by the means of raising the mobile robot intelligence level
and human machine interaction capabilities. This PhD thesis deals with first level problematic. Researched problems are related to odometry calibration, mobile
robot localization and map-building. Problems of mobile robot localization and
map-building are combined together in an implemented simultaneous localization
and map-building algorithm. In order to achieve faster development of algorithms, a mobile robot simulator for Matlab is developed. For a more realistic
simulation a possibility to set system and non-system error is added. The research subject firstly includes odometry calibration as a basic method of estimation
of mobile robot pose. This research part resulted also in a developed automatic
on-line odometric calibration system, which is expanded into an on-line odometric calibration system. Thereupon mobile robot localization is researched with
emphasis on local localization using two nonlinear Kalman filter version and a
local pose recovery method, which reduces implemented algorithm localization ti-
me. The third research subject includes occupancy grid map-building algorithms.
Today mostly used methods for on-line occupancy grid map-building are implemented and compared regarding their complexity and memory usage. Research in this thesis includes a simultaneous localization and map-building algorithm based
on a feature map. Ultrasound sensors placed on the mobile robot body are used
as the prime perception sensors in this thesis. All algorithms are implemented and tested both in simulations, using the developed AMORsim simulator, and
experimentally, using the mobile robots Pioneer 2 DX i Pioneer 3 DX.
BibTeX entry:
@phdthesis \{Ivanjko2009_484,
author = \{Ivanjko, E.},
title = \{Autonomous Navigation of Mobile Robots Based on Ultrasound Range Sesnsors},
school = \{Faculty of Electrical Engineering and Computing, University of Zagreb, Croatia},
year = \{2009}
No file available for download






About Us