Coordinated Motion Control of Swarms with Dynamic Connectivity in Potential Flows

This paper presents a general framework for coordinated motion control of autonomous swarms in the presence of obstacles. The proposed framework judiciously combines concepts and techniques from potential flows, artificial potentials and dynamic connectivity to realize complex swarm behaviors. To begin with, existing concepts from potential flows in fluid mechanics are used to solve the single-agent navigation problem. As an extension, an analytical solution to the stagnation point problem is provided. The potential flow based framework is then modified significantly to facilitate the coordinated control of swarms navigating through multiple obstacles. Artificial potentials are employed for swarming as well as enhanced obstacle avoidance. A novel concept of dynamic connectivity is utilized to improve the performance of obstacle avoidance (Line of Sight Connectivity) and to organize diverse swarm behaviors (Random Connectivity). Simulation results with a set of developed algorithms are included to illustrate the viability the proposed framework.