Mobile robots:localization and mapping alogorithms

Abstract

In recent years, mobile robots start to be very often used in various applications involving home automation, planetary exploration, regions surveillance, rescue missions, ruins exploration. In all these elds, to accomplish its tasks, a mobile robot needs to navigate into an environment facing the localization, mapping and path planning problems. In this thesis a set of new algorithms to solve the mobile robots localization and mapping problems are proposed. The aim is to provide an accurate mobile robot estimated pose and to build a reliable map for the environment where the robot moves, ensuring algorithms computational costs low enough so that the algorithms can be used in real time, while the robot is moving. The localization problem is faced in static and dynamic contexts, assuming the environment surrounding the robot completely known, partially known or totally unknown. In the static context, localization algorithms based on the use of cameras and inertial measurement units are proposed. In the dynamic context, the mobile robots localization problem is solved by developing a set of new Kalman lter versions. About the mapping problem, two novel mapping models are de- ned. These models are used along with the proposed localization techniques to develop three algorithms able to solve the Simultaneous Localization and Mapping (SLAM) problem. All the proposed solutions are tested through numerical simulations and experimental tests in a real environment and using a real mobile robot. The results show the e ectiveness of the proposed algorithms, encouraging further researches.