Formation Control of Ground Multi-agent System Using Quadcopter with Camera

The paper is dedicated to discussing a method for ground multi-agent formation control in different obstacle environments. Use the data from the quadcopter vision system and the lidar sensor of the ground agent to create a global map. Then, we perform coordinate estimation of the ground agent by using the terrestrial agent’s odometer, GPS, and quadcopter vision system. Finally, we use the finite state machine theory to control the switching of ground multi-agent formation, and propose a strategy to avoid ground multi-agent collisions when the formation is switched. The effectiveness of this strategy is demonstrated through simulation experiments.