A highly flexible virtual physics-based control framework for multi-robot chemical plume tracing

Abstract

This paper presents a virtual-physics based control framework for swarm robotic chemical plume tracing and source localization problem. The control force includes three kinds of effort, which are lattice formation force, plume tracing force, and obstacle avoidant force. The plume tracing and source identify strategy are based on the chemical mass flux passing through the robot colony. Simulation results show that the proposed control framework and tracing strategy can effectively navigate the multi-robot system to the chemical source emitter region. The virtual-physics based framework is highly flexible, endows the robot formation the properties of self-organization and self-repair. The mass flux driven plume tracing can inhibit the robots from adhering to plume "trap" and guide the robots forward to the real source region.