EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Deep Reinforcement Learning for Collision-Aware Leader–Follower Formation Control

Sławomir Romaniuk iD, Jakub Budnik

Sensors · 01.07.2026 · 26 (13) · 4162

DOI: 10.3390/s26134162 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 01.07.2026. Verfasst von Sławomir Romaniuk und 1 weiteren Autor:innen. Beteiligte Einrichtung: Department of Automatic Control and Robotics, Faculty of Electrical Engineering, Bialystok University of Technology, 15-351 Bialystok, Poland. Die Arbeit nennt 31 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Bialystok University of Technology.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This work tackles leader–follower navigation in scenarios where rapid, unpredictable leader maneuvers can trigger unsafe proximity or collisions. We apply Proximal Policy Optimization (PPO) to learn adaptive follower behavior capable of tracking the leader’s direction while considering a minimum-distance safety objective. A task-specific reward function—capturing formation constraints, collision avoidance, control-effort regularization, and responsiveness to sudden directional changes—enables effective policy learning. The results show that the PPO-based follower provides a favorable safety–tracking trade-off compared with the considered A2C and PID baselines under the simulated test conditions, although collision-free operation is not guaranteed in all episodes.

Schlagwörter

Stichwörter der Autor:innen

MuJoCoautonomous robotscollision avoidancedeep reinforcement learningleader–follower formationmobile roboticsproximal policy optimization

Autor:innen und Institutionen

  • Department of Automatic Control and Robotics, Faculty of Electrical Engineering, Bialystok University of Technology, 15-351 Bialystok, Poland.

ORCID-Profile: Sławomir Romaniuk

Förderung

  • Bialystok University of Technology

Zitieren

APA

Romaniuk, S., & Budnik, J. (2026). Deep Reinforcement Learning for Collision-Aware Leader–Follower Formation Control. Sensors, 26(13), 4162. https://doi.org/10.3390/s26134162

Vancouver

Romaniuk S, Budnik J. Deep Reinforcement Learning for Collision-Aware Leader–Follower Formation Control. Sensors. 2026;26(13):4162. doi:10.3390/s26134162

BibTeX
@article{Romaniuk2026,
  author = {Romaniuk, Sławomir and Budnik, Jakub},
  title = {Deep Reinforcement Learning for Collision-Aware Leader–Follower Formation Control},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {13},
  pages = {4162},
  doi = {10.3390/s26134162}
}
Kostenlos starten →