.png){kind=icon}
.png){kind=icon}
Return
Reinforcement learning-based event-triggered fuzzy control for unmanned surface vehicles under malicious attacks
.png){kind=icon}
PRE
AI
H
X
J
L
DOI：10.1016/j.fss.2026.110030.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This article concentrates on the reinforcement learning-based event-triggered fuzzy control problem for unmanned surface vehicle (USV) systems subject to aperiodic denial-of-service (DoS) attacks. In view of the nonlinearity and variability of the marine environment, the USV systems are characterized by a Takagi-Sugeno (T-S) fuzzy model. A switched event-triggered protocol is proposed to ensure the optimal utilisation of communication resources and compensate the data interruptions caused by DoS attacks. Then, to obtain the Nash equilibrium solution of the zero-sum game problem, a data-driven composite policy iteration algorithm based on the experience replay technique is developed, adaptively learning the optimal control policies from system data, which weakens the condition for initial stabilizing control policies and accelerates the learning process compared to traditional online policy iteration and value iteration algorithms. The stability and convergence of the proposed algorithm are rigorously demonstrated. Finally, the effectiveness and good performance of the proposed control method are validated through extensive simulations and experiments.
Journal
IF:
2.7
Papers:
7.6K
Citations:
1.5W
Organization
Cited Papers