Social Insect-Inspired Multi-Robot Coverage

摘要

Coordination is one of the most interesting and complicated research issues in distributed multi-robot systems (MRS), aiming to improve performance, energy consumption, robustness and reliability of a robotic system in accomplishing complex tasks. Social insect-inspired coordination techniques achieve these goals by applying simple but effective heuristics from which elegant solutions emerge. This paper demonstrates a hybrid ant-and-bee inspired approach, HybaCo, designed to provide coordinated multi-robot solutions to area coverage problems. We explore performance consequences of ant-inspired (StiCo) and beeinspired (BeePCo) approaches, first separately and then compared to each other according to a number of features. The proposed hybrid approach is evaluated in multiple scenarios using a high-level simulator. Experimental results from various scenarios indicate that HybaCo improves the area coverage uniformly. The contribution of this work lies not only in the studied comparison and combination of two methods from the literature, but also in the methodology of evaluation and proposal of innovative metrics that capture differences in the feature spaces of the methods considered.