Laser sensor based localization of mobile robot using Unscented Kalman Filter

Abstract

The objective is to determine mobile robots position and orientation by integrating information received from laser distance sensor and encoders. The robot is maneuvered in a known environment, and the laser ranging finder can get information of geometrical primitives like lines and polygons to extract landmarks of the environment. With the off-line map, the position and orientation of the robot can be estimated. To improve the precision of our localization system, we present a sensor-data-fusion method using Unscented Kalman Filter (UKF).