Multi-robot Autonomous Cooperative Exploration in Unknown Environment Based on Region Assignment

Abstract

To address the low efficiency of multi-robot exploration, this paper proposes a novel method based on region assignment to guide multi-robot teams to select the frontier points with the highest gain each time. Additionally, a frontier enclosure function is designed to prevent multiple robots from converging towards the same or adjacent frontier points, thereby enabling the team to disperse and explore collaboratively across the entire unknown environment. The objective of this method is to minimize the total exploration time required by a multi-robot team to complete the exploration of a specific unknown area, while the team utilizes onboard Lidar sensors to create a global map. Simulation results demonstrate that our method outperforms the recent advanced open-source approach.