Enhancing drone swarm efficiency through a high-flexibility biomimetic formation algorithm

With the rapid advancement of unmanned aerial vehicle technologies, drone swarms have been increasingly adopted in applications ranging from agriculture and logistics to defense and performance arts. However, conventional swarm control architectures, predominantly centralized, remain limited in scalability, adaptability, and robustness under dynamic conditions. To address these limitations, this study presents a bio-inspired formation control framework that integrates decentralized coordination and autonomous role assignment. The proposed system incorporates a reference–follower mechanism, enabling drones to dynamically select reference units based on spatial proximity, thereby enhancing inter-drone interaction and formation stability. Furthermore, a hybrid communication architecture based on robot operating system (ROS) and message queuing telemetry transport protocols is developed to overcome the constraints of traditional ROS communication frameworks and improve scalability. The proposed framework shows strong potential for application in various domains such as precision agriculture, search and rescue, and environmental monitoring, offering a flexible and adaptive solution for future drone swarm operations. Finally, simulation and real-world flight experiments validate the proposed approach, demonstrating significant improvements in formation flexibility, communication efficiency, and adaptive leader switching compared to conventional leader–follower and virtual structure models.