Real-time map building and navigation for autonomous robots inunknown environments

摘要

An algorithmic solution method is presented for the problem ofnautonomous robot motion in completely unknown environments. Our approachnis based on the alternate execution of two fundamental processes: mapnbuilding and navigation. In the former, range measures are collectednthrough the robot exteroceptive sensors and processed in order to buildna local representation of the surrounding area. This representation isnthen integrated in the global map so far reconstructed by filtering outninsufficient or conflicting information. In the navigation phase, annA*-based planner generates a local path from the current robot positionnto the goal. Such a path is safe inside the explored area and provides andirection for further exploration. The robot follows the path up to thenboundary of the explored area, terminating its motion if unexpectednobstacles are encountered. The most peculiar aspects of our method arenthe use of fuzzy logic for the efficient building and modification ofnthe environment map, and the iterative application of A*, a completenplanning algorithm which takes full advantage of local information.nExperimental results for a NOMAD 200 mobile robot show the real-timenperformance of the proposed method, both in static and moderatelyndynamic environments