Autonomous Exploration Using a Tree Structure For Goal Selection
Barbara Abonyi-Tóth, Ákos Nagy · 2023
Autonomous exploration of unknown environments using unmanned robots is a widely researched problem of our days. The focus of this paper is to provide a novel goal-selection method in response to the problems of the simple greedy goal selection algorithm. The presented method uses a tree structure built from the detected frontiers and selects the next goal of exploration using a depth-first search on the expanding tree. This prevents the exploring robot from leaving half-explored areas and backtracking to them later in the exploration. The method is tested in two simulated environments. The results are compared to those achieved using a simple greedy goal selection algorithm.