Mean-Field-Based Density Control for Swarm Robotics Passing-Through a Virtual Tube

Navigating large swarms of robots through obstacle-dense environments has attracted increasing attention. Current studies have mainly adopted a bottom-up approach, relying on local information and control laws. However, analyzing the global behaviors of the swarm, such as the density distribution, remains a significant challenge. This letter simplifies navigation in cluttered environments to passing through a virtual tube with varying widths. A novel approach integrating the virtual tube concept with mean-field-based density control is proposed to enhance safety and efficiency, guiding the swarm toward the desired configuration. The key contribution is the integration of macroscopic behaviors and control within a virtual tube, providing an effective solution to the swarm passing-through problem.