Control algorithms for the emergence of self-organized behaviours in swarms of differential-traction wheeled mobile robots

摘要

This article proposes three control algorithms for the emergence of self-organized behaviours, including aggregation, flocking and rendezvous, in swarm robotics systems. The proposed control algorithms are based on a local polar coordinates’ control law available in the literature for posture regulation; this law is adapted to work in a self-organized robotic swarm using distance and bearing as coupling information. Therefore, the robots only need to know the radial distance and orientation to the goal; additionally, the three algorithms are based on self-organization, eliminating the need for a preset coupling topology among the robots. In particular, the flocking algorithm has a first stage for topology creation, while the rendezvous and aggregation algorithms change the topology on every iteration depending on the local interactions of the robots. The effectiveness of the algorithms was evaluated through numerical simulations of swarms of up to 100 differential traction wheeled mobile robots.