UNSCENTED KALMAN FILTER POSITION ESTIMATION FOR AN AUTONOMOUS MOBILE ROBOT

Abstract

The Kalman filters have been widely used for mobile robot navigation and system integration. So that it may operate autonomously, a mobile robot must know where it is. Accurate localization is a key prerequisite for successful navigation in large-scale environments, particularly when global models are used, such as maps, drawings, topological descriptions, and CAD models. This paper presents the localization of a mobile robot using one variation of the traditional Kalman filter: the unscented Kalman filter (UKF). For this purpose the filter was implemented for a known kinematic model of the robot.