A method for position and orientation determination of a mobile robot equipped with laser scanner

This paper presents an algorithm for determination unknown position and orientation of a mobile robot indoors. The algorithm is based on a selected set of features for landmarks in robot's environment. Initial data are comprised of horizontal distance measurements performed by the robot's laser scanner. The data is used to create a current state model for the visible surroundings of the robot, from wich metric characteristics of the features are extracted by a special procedure. Robot's control system database contains a map of the environment with features extracted by a similar procedure. Within this map a zone is chosen which closely matches a set of features for the current state model. Robot's position and orientation are then evaluated based on the relative position of pairs of similar features. Experiments show that proposed algorithm is fast enough to evaluate robot's position with sufficient precision to start movement upon turning on the robot.