A multi-robot cooperative exploration algorithm considering working efficiency and working load

Abstract

The cooperative exploration in unknown environment is a tough task for the multi-robot system. The imbalance of individual workload caused by the weak autonomous cooperation ability will affect the working efficiency of the multi-robot system. In this paper, a two-objective cooperative exploration algorithm (TOCEA) is proposed, where the working efficiency and working load at the individual and system levels are both considered. There are three parts in the proposed algorithm, namely, frontier detection, target point selection and exploration decision. First, based on the analysis of the frontier characteristics, the accumulation of local frontiers is used to replace the method of traversal search, which greatly improves the exploration efficiency. Second, each robot plays as an equal role to select their candidate target points from the individual level, and uses the maximum area criterion to select the final target points. Finally, the robot target points are clustered to reduce the repeated exploration of the robots, which is essential for reduce the path length and exploration time. Significantly, the whole exploration process is completely based on the autonomous cooperation of the robots. The experimental results performed on three different platforms illustrate that the TOCEA shows improvements in terms of working efficiency, cooperation performance and applicable fields.