Simultaneous Localization and Dynamic State Estimation in Reconfigurable Environments

The majority of existing approaches to mobile robot localization assume that the world is static, which clearly does not hold in most real-world application domains. In this paper we present a probabilistic approach to global localization in reconfigurable environments, where the robot pose and the environment state are jointly estimated using a RaoBlackwellized particle filter. The environment is represented as a spatial grid and a hidden Markov model is used to represent the occupancy state and state transition probabilities of each grid cell. The HMM parameters are estimated offline using the EM algorithm. Experimental results show that our model is better suited for representing reconfigurable environments than standard occupancy grids. Furthermore, the results show that the explicit representation of the environment dynamics can be used to improve localization accuracy in reconfigurable environments.