EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

MW-MADDPG: a meta-learning based decision-making method for collaborative UAV swarm

Minrui Zhao, Gang Wang, Qiang Fu, Xiangke Guo, Yu Chen, Tengda Li, XiangYu Liu

Frontiers in Neurorobotics · 21.09.2023 · 17

DOI: 10.3389/fnbot.2023.1243174 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Frontiers in Neurorobotics, online erschienen am 21.09.2023. Verfasst von Minrui Zhao und 6 weiteren Autor:innen. Beteiligte Einrichtung: College of Air and Missile Defense, Air Force Engineering University, Xi'an, China. Die Arbeit wurde bisher 28-mal zitiert (Crossref) und nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Unmanned Aerial Vehicles (UAVs) have gained popularity due to their low lifecycle cost and minimal human risk, resulting in their widespread use in recent years. In the UAV swarm cooperative decision domain, multi-agent deep reinforcement learning has significant potential. However, current approaches are challenged by the multivariate mission environment and mission time constraints. In light of this, the present study proposes a meta-learning based multi-agent deep reinforcement learning approach that provides a viable solution to this problem. This paper presents an improved MAML-based multi-agent deep deterministic policy gradient (MADDPG) algorithm that achieves an unbiased initialization network by automatically assigning weights to meta-learning trajectories. In addition, a Reward-TD prioritized experience replay technique is introduced, which takes into account immediate reward and TD-error to improve the resilience and sample utilization of the algorithm. Experiment results show that the proposed approach effectively accomplishes the task in the new scenario, with significantly improved task success rate, average reward, and robustness compared to existing methods.

Schlagwörter

Stichwörter der Autor:innen

MADDPGModel Agnostic Meta Learning (MAML)UAVmeta learningmulti-agent reinforcement learning (MARL)

Autor:innen und Institutionen

  • College of Air and Missile Defense, Air Force Engineering University, Xi'an, China.
  • Graduate School, Academy of Military Science, Beijing, China.
  • Unit 95866 of PLA, Baoding, China.

Zitieren

APA

Zhao, M., Wang, G., Fu, Q., Guo, X., Chen, Y., Li, T., & Liu, X. (2023). MW-MADDPG: a meta-learning based decision-making method for collaborative UAV swarm. Frontiers in Neurorobotics, 17. https://doi.org/10.3389/fnbot.2023.1243174

Vancouver

Zhao M, Wang G, Fu Q, Guo X, Chen Y, Li T, et al. MW-MADDPG: a meta-learning based decision-making method for collaborative UAV swarm. Frontiers in Neurorobotics. 2023;17. doi:10.3389/fnbot.2023.1243174

BibTeX
@article{Zhao2023,
  author = {Zhao, Minrui and Wang, Gang and Fu, Qiang and Guo, Xiangke and Chen, Yu and Li, Tengda and Liu, XiangYu},
  title = {MW-MADDPG: a meta-learning based decision-making method for collaborative UAV swarm},
  journal = {Frontiers in Neurorobotics},
  year = {2023},
  volume = {17},
  doi = {10.3389/fnbot.2023.1243174}
}
Kostenlos starten →