Vision-Based Formation Control for a Heterogeneous Multi-Robot System

This work presents a decentralized formation control approach for a heterogeneous multi-robot system including a team of nonholonomic unmanned ground vehicles and an unmanned aerial vehicle based on visual localization. Relying only on the visual information from the onboard camera mounted on the unmanned aerial vehicle, the unmanned ground vehicle team can be controlled in a desired formation with the flying drone. A gradient-based control law is designed, in which the formation producing, collision avoidance and field of view constraint have been considered comprehensively. Finally, the performance and validness of the developed method are verified through simulations.