Decentralized control for swarm flocking in 3D space

Abstract

This paper presents a decentralized control strategy for a robot swarm where each robot tries to form a regular tetrahedron with its three neighbors. The proposed method is based on virtual spring. Robots can form regular tetrahedron regardless of their initial positions and they require minimum amount of information about their neighbors. The control strategy is made scalable by integrating a neighbor selection procedure so that it can be expanded to large swarms easily. In addition, an obstacle avoidance mechanism, based on artificial physics, is also introduced. By utilizing this control strategy, basic swarm behaviors such as aggregation, flocking and obstacle avoidance are demonstrated through simulations in an unknown three dimensional environment. © 2009 Springer-Verlag Berlin Heidelberg.