.png){kind=icon}
.png){kind=icon}
Return
Decentralized Self-Triggered Learning Control for Constrained Large-Scale Networked Systems
.png){kind=icon}
PRE
AI
DOI：10.1016/j.cnsns.2026.110803.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In this article, a reinforcement learning-integrated decentralized self-triggered control strategy is proposed for large-scale networked systems (LSNSs) in the presence of asymmetric input constraints. To alleviate the dual burden of communication and computation, a novel decentralized dynamic self-triggered mechanism (DDSTM) is developed to proactively determine the subsequent triggering instants with the Zeno-excluded guarantee. In contrast to conventional event-triggered control, the DDSTM no longer requires the embedded hardware to continuously monitor the triggering condition. Furthermore, a critic-sole neural network is utilized to approximate the non-quadratic optimal cost function with the integration of the asymmetric-constrained features and the upper bound of interconnection. Under the proposed scheme, the states of all auxiliary subsystems and the critic weight approximation errors are guaranteed to be uniformly ultimately bounded. Finally, two simulation examples, involving interconnected chemical reactors and an LSNS composed of one hundred subsystems, verify the effectiveness and practicability of the developed algorithm.
Journal
IF:
3.8
Papers:
9.2K
Citations:
1.8W
View more organizations