Set membership localization of mobile robots via angle measurements

This paper addresses the localization problem for a mobile robot navigating in an unstructured outdoor environment. A new technique is introduced, for computing an estimate of the position of the robot and the related uncerta inty region, in the presence of visual angle measurements affected by bound ed errors. The proposed set membership estimation procedure exploits the st ructure of the static set estimator, to solve recursively the dynamic local ization problem.