Extended Kalman Filter for indoor and outdoor localization of a wheeled mobile robot

Abstract

This paper presents a robot localization algorithm, that uses an Extended Kalman Filter (EKF) to fuse data from optical wheel encoders, a gyroscope and an accelerometer for an indoor navigation and additionally from DGPS unit for an outdoor scenario. The algorithm's performance is experimentally evaluated using a skid-steered SeekurJr mobile robot. Experimental results are provided to compare the localization accuracy achieved using the proposed algorithm with those using pure odometry readings and pure DGPS readings.