EUVIMEDEUVIMED entdecken
Crossref

Endogenous Cooperative Game for Guidance–Sensing Coupling in Multispacecraft Pursuit–Evasion

Gena Gan, Ming Chu, Huayu Zhang, Shaoqi Lin

IEEE Transactions on Aerospace and Electronic Systems · 2026 · 62 · 13664-13697

DOI: 10.1109/taes.2026.3710455 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Aerospace and Electronic Systems, 2026. Verfasst von Gena Gan und 3 weiteren Autor:innen. Die Arbeit nennt 43 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Beijing Municipal Natural Science Foundation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In angles-only multispacecraft cooperative pursuit–evasion, guidance maneuvers directly reshape the observation geometry, causing the traditional separation principle to fail. Prior approaches—information-aware model predictive control (MPC), differential game, and end-to-end reinforcement learning—leave this guidance–sensing coupling unresolved because none treats the two subsystems as endogenously coupled agents. We propose the endogenous cooperative game architecture (ECGA), which reformulates guidance and sensing as peer game players with explicitly coupled cost functions. A bidirectional mechanism driven by an information demand factor and an adaptation cost negotiates a Nash equilibrium that jointly optimizes interception effectiveness and observability maintenance. ECGA is realized by a proximal policy optimization guidance agent and an adaptive dynamic programming sensing agent coordinated through a confidence-weighted cooperative observation network, supporting offline centralized training and efficient onboard execution. Across 500 Monte Carlo runs, ECGA achieves a 91.5% interception success rate; against a capacity-matched decoupled baseline, peak estimation uncertainty is reduced by 87.7% and the Fisher information matrix condition number by 85.2%, confirming that these gains stem from the game coupling mechanism rather than network capacity. Compared to separation-based, team-optimal MPC, and external adaptive dynamic programming baselines, ECGA offers Pareto-optimal advantages in interception success, observability maintenance, and fuel economy, while reducing mean capture time by 19.7% under the same controlled ablation.

Förderung

  • National Natural Science Foundation of China
  • Beijing Municipal Natural Science Foundation

Zitieren

APA

Gan, G., Chu, M., Zhang, H., & Lin, S. (2026). Endogenous Cooperative Game for Guidance–Sensing Coupling in Multispacecraft Pursuit–Evasion. IEEE Transactions on Aerospace and Electronic Systems, 62, 13664-13697. https://doi.org/10.1109/taes.2026.3710455

Vancouver

Gan G, Chu M, Zhang H, Lin S. Endogenous Cooperative Game for Guidance–Sensing Coupling in Multispacecraft Pursuit–Evasion. IEEE Transactions on Aerospace and Electronic Systems. 2026;62:13664-13697. doi:10.1109/taes.2026.3710455

BibTeX
@article{Gan2026,
  author = {Gan, Gena and Chu, Ming and Zhang, Huayu and Lin, Shaoqi},
  title = {Endogenous Cooperative Game for Guidance–Sensing Coupling in Multispacecraft Pursuit–Evasion},
  journal = {IEEE Transactions on Aerospace and Electronic Systems},
  year = {2026},
  volume = {62},
  pages = {13664--13697},
  doi = {10.1109/taes.2026.3710455}
}
Kostenlos starten →