COMPARISON OF EXPLORATION STRATEGIES FOR MULTI-ROBOT SEARCH

Acta Polytechnica

Libor Přeučil¹  Miroslav Kulich¹  Tomáš Juchelka¹ 

[1] Department of Cybernetics, Faculty of Electrical Engineering, Czech Technical University in Prague, Technicka2, 166 27 Prague, Czech Republic

关键词: mobile robots;    multi-robot systems;    planning;    search;    exploration;   

DOI  :  

来源: Czech Technical University in Prague, Faculty of M

【 摘 要 】

Searching for a stationary object in an unknown environment can be formulated as an iterative procedure consisting of map updating, selection of a next goal and navigation to this goal. It finishes when the object of interest is found. This formulation and a general search structure is similar to the related exploration problem. The only difference is in goal-selection, as search and exploration objectives are not the same. Although search is a key task in many search and rescue scenarios, the robotics community has paid little attention to the problem. There is no goal-selection strategy that has been designed specifically for search. In this paper, we study four state-of-the-art strategies for multi-robot exploration, and we evaluate their performance in various environments with respect to the expected time needed to find an object, i.e. to achieve the objective of the search.

【 授权许可】

Unknown