Comparison of Mapmaking Methods for Mobile Robots

The problem of creating the map of an unknown environment by mobile robot control system is considered using range readings obtained by ultrasonic range finder. A grid-based representation of a robot working space is chosen. The goal is the determination of occupied and empty areas in the working environment, which is not a simple problem due the uncertainty introduced by the sensory system. Three different algorithms for map generation based on probability theory, the Dempster– Shafer theory of evidence and fuzzy set theory are compared. Experimental examples of occupancy grids, built from real data recorded in an office-like environment, are presented.