An Effective Kalman Filter Localization Method for Mobile Robots

Abstract

An effective Kalman filter localization method for mobile robots is investigated in terms of the robust Kalman filter with perturbation estimator. In the recursive algorithm, the perturbation estimator produces equivalent perturbations with respect to the nominal state equation and the action model of a mobile robot is adaptively modified with the perturbation estimates. The integral control property of the perturbation estimator enables a great reduction of the localization error, specifically when the odometric disturbance is large. The Kalman filter recursive equations including predictor, corrector, perturbation estimator, and the corresponding covariance propagation equations are formulated systematically. The effectiveness of the proposed scheme is verified through simulation and experimental results for a wheeled mobile robot