Game-Theoretic Fault-Tolerant Strategies for Multi-Agent Pursuit-Evasion With Actuator Faults
IEEE Control Systems Letters · 2026 · 10 · 217-222
DOI: 10.1109/lcsys.2026.3683403 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Control Systems Letters, 2026. Verfasst von Wenjing Hou und 2 weiteren Autor:innen. Die Arbeit nennt 21 Literaturangaben. Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
This letter studies game-theoretic fault-tolerant strategy design for multi-agent pursuit-evasion (MPE) systems subject to actuator faults. By explicitly embedding actuator fault effects into the strategy layer, the MPE interaction is formulated as a zero-sum dynamic game in which the pursuers aim to minimize capture errors in the presence of both evader actions and fault-induced uncertainties. To address the associated Hamilton-Jacobi-Isaacs (HJI) equation, a critic-only online adaptive dynamic programming algorithm is developed, where a single critic neural network approximates the optimal value function and corresponding feedback strategy via experience replay. The proposed approach yields an implementable fault-tolerant strategy with reduced computational complexity. Simulation results verify the effectiveness and robustness of the proposed method under actuator fault scenarios.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Hou, W., Liang, L., & Wang, Y. (2026). Game-Theoretic Fault-Tolerant Strategies for Multi-Agent Pursuit-Evasion With Actuator Faults. IEEE Control Systems Letters, 10, 217-222. https://doi.org/10.1109/lcsys.2026.3683403
Vancouver
Hou W, Liang L, Wang Y. Game-Theoretic Fault-Tolerant Strategies for Multi-Agent Pursuit-Evasion With Actuator Faults. IEEE Control Systems Letters. 2026;10:217-222. doi:10.1109/lcsys.2026.3683403
BibTeX
@article{Hou2026,
author = {Hou, Wenjing and Liang, Li and Wang, Youqing},
title = {Game-Theoretic Fault-Tolerant Strategies for Multi-Agent Pursuit-Evasion With Actuator Faults},
journal = {IEEE Control Systems Letters},
year = {2026},
volume = {10},
pages = {217--222},
doi = {10.1109/lcsys.2026.3683403}
}
EUVIMED