Improving Formation in Swarm Robotics with a Leader-Follower Approach

Abstract

Collective Motion (CM) is a bio-inspired swarm behaviour observed in nature, such as schooling fish and flocking birds. In swarm robotics, CM is attained through the implementation of virtual viscoelastic links that generate attractive and repulsive forces among robots. This study investigates the challenges associated with maintaining formation integrity in specific geometrical arrangements, such as ring configurations. To address this, a Leader-Follower (L-F) approach is proposed to enhance the swarm shape stability. Simulation results demonstrated that the proposed method significantly optimised CM performance, achieving alignment of nearly 99% of the robots within 50 s. This shows the effectiveness of the L-F approach in successfully implementing swarm formation with a significant performance improvement.