Distributed Model Predictive Control with High-Order Control Barrier Functions for Swarm Robotics within a 3D Virtual Tube

Achieving safe and efficient collaborative navigation for swarm robotics in complex, unknown environments filled with dynamic obstacles remains a critical challenge and research focus in the field of unmanned aerial vehicles. This paper proposes an innovative framework that employs the Rapidly-exploring Random Tree Star (RRT*) algorithm to generate a virtual tube that is free of obstacles and integrates model predictive control (MPC) with a high-order control barrier function (HOCBF) to establish a unified and robust control system. This system enhances the ability of the robot swarm as a whole to navigate safely to its destination. This approach not only enables the swarm to stably navigate to the target region but also provides a practical solution for collision avoidance in obstacle-dense environments.