Robust mobile robot localization with combined Kalman filter-perturbation estimator

In this paper, a robust localization method for mobile robot based on the combination of Kalman filter and perturbation estimator is presented. It remarkably enhances the robustness of localization performance, specifically when large odometric errors are occurred. The perturbation estimator in the combined Kalman filter (CKF) is to estimate systematic errors which perturbs the behavior of nominal state transition equation. Intrinsically, it has the property of integrating the innovation, i.e., the difference between measurement and predicted measurement and thus gives a chance of more reducing the gap between real states and their estimates. After formulation of the CKF recursion, we show how the design parameters can be determined and how much beneficial it is through simulation and experiment for a two-wheeled mobile robot under indoor GPS.