Accurate mobile robot position determination using unscented Kalman filter

摘要

The objective is to accurately determine mobile robots position and orientation by integrating information received from odometry and an inertial sensor. The position and orientation provided by odometry are subject to different types of errors. To improve the odometry, a fiber optic gyroscope is used to give the orientation information that is more reliable. The information from odometry and gyroscope are integrated using unscented Kalman filter (UKF). The position and orientation determined using the UKF is compared with the results from the extended Kalman filter (EKF)