Probabilistic Grid Map Based Localizability Estimation for Mobile Robots

Abstract

Based on the widely used probabilistic grid map, a localizability estimation method for mobile robots is pro- posed. Firstly, the Fisher information matrix (FIM) of robot localization is transformed into discrete form, and a static localizability matrix suitable for off-line estimation based on the known grid map is obtained. On this basis, the impact factor of locally sensed unkown obstacles is adopted to modify the static localizability matrix, and a dynamic localizability matrix is proposed for on-line estimation to deal with unexpected dynamic changes of environments. This matrix describes both the localizability index and localizability direction of mobile robots quantitatively. The results of real robot experiments under different typical environments demonstrate the validity of the proposed method.