EUVIMEDEUVIMED entdecken
Crossref

Learning-Based Adaptive Morphing Flight Control for Disturbance Resistance of Morphing Quadrotor UAVs via Meta Reinforcement Learning

Huai-Ning Wu, Jin-Feng Zhang, Tao Yang, Jun-Wei Wang

IEEE Transactions on Aerospace and Electronic Systems · 2026 · 62 · 19344-19356

DOI: 10.1109/taes.2026.3735975 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Aerospace and Electronic Systems, 2026. Verfasst von Huai-Ning Wu und 3 weiteren Autor:innen. Die Arbeit nennt 61 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Institute of Beihang University.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Inspired by the bird’s flight, practical applications have given rise to a concept of morphing unmanned aerial vehicles (UAVs) recently. This type of UAV can actively change its shape or structure to maintain optimal performance in disturbance environments. However, because the relationship between the shape or structure change of morphing UAVs and its effect on aircraft maneuverability is not well understood, it is challenging to design effective flight controllers using traditional control design methods. This study presents a Reptile-proximal policy optimization (Reptile-PPO) algorithm for disturbance resistance of a morphing quadrotor UAV (MQ-UAV) with active variable-length arm rods. This is a learning-based adaptive morphing flight control policy using a small amount of flight data that enables rapid online adaptation for the MQ-UAV in disturbance environments by incorporating the PPO algorithm through meta-learning. In the proposed morphing flight control policy, the PPO algorithm and reptile one are assigned as a base-learner and the meta-learner, respectively. Both a long-term indicator and a short-term one are introduced to evaluate the morphing flight performance when the proposed morphing flight control policy is deployed to unseen disturbance environments. Simulation results are provided to support the theoretical algorithm.

Förderung

  • National Natural Science Foundation of China
  • Institute of Beihang University

Zitieren

APA

Wu, H., Zhang, J., Yang, T., & Wang, J. (2026). Learning-Based Adaptive Morphing Flight Control for Disturbance Resistance of Morphing Quadrotor UAVs via Meta Reinforcement Learning. IEEE Transactions on Aerospace and Electronic Systems, 62, 19344-19356. https://doi.org/10.1109/taes.2026.3735975

Vancouver

Wu H, Zhang J, Yang T, Wang J. Learning-Based Adaptive Morphing Flight Control for Disturbance Resistance of Morphing Quadrotor UAVs via Meta Reinforcement Learning. IEEE Transactions on Aerospace and Electronic Systems. 2026;62:19344-19356. doi:10.1109/taes.2026.3735975

BibTeX
@article{Wu2026,
  author = {Wu, Huai-Ning and Zhang, Jin-Feng and Yang, Tao and Wang, Jun-Wei},
  title = {Learning-Based Adaptive Morphing Flight Control for Disturbance Resistance of Morphing Quadrotor UAVs via Meta Reinforcement Learning},
  journal = {IEEE Transactions on Aerospace and Electronic Systems},
  year = {2026},
  volume = {62},
  pages = {19344--19356},
  doi = {10.1109/taes.2026.3735975}
}
Kostenlos starten →