Reinforcement Learning-Based UAV Swarm Fission–Fusion Approach With Real-World Data-Integrated Validation
Mar 28, 2025·
,,,·
1 min read
Xiaorong Zhang
Equal contribution
Dacheng Qi
Equal contribution
,Wenrui Ding
Xinrui Zhang
Qingyi Liu
Yufeng Wang
Image credit: UnsplashAbstract
This paper proposes a reinforcement learning–based UAV swarm fission–fusion approach with real-world data integrated validation for the swarm’s fission-fusion behavior in response to multiple unknown dynamic disturbances, along with a system validation method utilizing real-world data. The proposed approach effectively integrates fission–fusion dynamics with perception and control to enable UAV swarms to function robustly in the presence of such disturbances. First, we develop a self-organized control framework that facilitates the coordinated motion of multiple UAV swarms. Second, we introduce a reinforcement learning–based fission–fusion confrontation algorithm designed to minimize resource consumption while effectively responding to multiple unknown dynamic disturbances. Finally, we present a real-world data-based validation system based on AirSim, which allows comprehensive evaluation of UAV swarm performance in actual environments.
Type
Publication
International Journal of Aerospace Engineering, 2025(1)
Please visit the official IJAE page for details.
Authors
Dacheng Qi
(he/him)
PhD Student
I am a Ph.D. student in Computing & Data Science at The University of Hong Kong, supervised by Prof. Yi Ma and Prof. Shenghua Gao. I received my M.S. degree from Beihang University under the supervision of Prof. Wenrui Ding and Researcher Yufeng Wang.
My research interests lie in generative methods for 3D design and manufacturing, especially parametric CAD synthesis and its integration with multimodal large language models. Before this, I worked on 3D generation.