Local map-based exploration using a breadth-first search algorithm for mobile robots

Abstract

This paper describes a local map-based exploration strategy. Segmented frontiers and relative transformations constitute a tree structure. This frontier tree structure, which manages multiple local maps, effectively overcomes the limitations of conventional exploration methods, which maintain only a single global map. Although this method uses only local maps and adjacent node information, mapping completion and efficiency can be improved greatly by merging and updating the frontier nodes. In addition, a modified breadth-first search (BFS) algorithm is used to determine the next exploration target. BFS exploration is appropriate for large environments because it induces a loop-closing event from the root node, which is necessary to recover the estimation accuracy when the uncertainty of the robot's pose has become large. The proposed local map-based BFS exploration can construct an accurate map, even in large environments.