Collective Navigation of Aerial Vehicle Swarms: A Flocking Inspired Approach
Fakui Wang, Jialei Huang, Kin Huat Low, Tianjiang Hu
Abstract
Collective navigation is inspired by natural migratory species traveling long distances and conserving their group behavior in the whole journey. With the technological development of autonomous intelligent systems, collective navigation has been an important application in the field of intelligent vehicles. However, many of the existing collective behavior approaches rely on communication exchanges between individuals and focus on the emergence of collective motion, which limits the application in specific mission scenarios and lacks robustness against interference. To address these challenges, this article proposes a novel flocking-inspired approach of collective navigation for aerial vehicle swarms flying synchronously to the desired region. Firstly, flocking without communication is achieved based on bionic visual projection field (VPF), ensuring the normal operation of the swarm in communication-denied unknown environments. Secondly, each aerial entity is assigned a hidden identity as informed or uninformed, but only the informed agents are allocated information about the migration destination. Under such an architecture, the trade-off between informed and uninformed individuals in the group enables few-to-many control and fast navigation. Finally, simulation and experiments are both conducted within multiple scenarios, in terms of adaptability to different environments, swarm scalability, and robustness to vehicle failures. Results validate the effectiveness of the developed bio-inspired collective navigation algorithm.