Unmanned ground vehicle localization by dead-reckoning/GPS sensor fusion

In autonomous navigation systems, localization is the most important problem to be solved. In this note, we propose the algorithm to localize an unmanned ground vehicle (UGV) equipped with a global positioning system (GPS), an odometer and an electronic compass. The proposed scheme is based on the unscented Kalman filter (UKF). To improve the accuracy and robustness, the novel noise model of data from an electronic compass is incorporated in the system modeling. The performance of our simple algorithm is evaluated by the experimental results that strongly confirm the benefit even during GPS outage in some short period of time.