UAV Swarm Formation Control for Outdoor Surveillance Tasks

Abstract

Multi-robot and swarm control is a promising field in modern robotics. This paper considers a formation control of an unmanned aerial vehicles’ (UAV) swarm focusing on implementation of a leader-follower algorithm within the Robot Operating System (ROS) framework, which is designed as a ROS plugin. The paper describes a concept of a swarm formation control and presents a set of typical formations that could be helpful for swarms’ tasks. The leader-follower algorithm was implemented and validated in the Gazebo simulator. The proposed algorithm novelty lies in its simplicity and effectiveness for outdoor applications, integration of collision-avoidance mechanisms of swarms’ UAVs, and synchronization of UAVs’ actions within the swarm. Virtual experiments with a small swarm of PX4-LIRS UAVs demonstrated consistent results in self-collision avoidance by the swarms’ members while performing formation generation and its further motion within outdoor areas. Expected and real coordinates of each UAV of the swarm were compared, and the control strategy achieved acceptably small absolute errors in UAVs’ positioning.