Reinforcement learning-based near-optimal control for discrete time-delay singularly perturbed systems with unmeasurable states

Meng Xu, Wei Dai, Qirui Zhang and Chunyu Yang

International Journal of Systems Science, 2025, vol. 56, issue 11, 2664-2673

Abstract: This brief proposes a novel reinforcement learning-based composite near-optimal output feedback (OPFB) control method for discrete time-delay singularly perturbed systems (SPSs) with unknown dynamics and unmeasured system states. Firstly, an optimal control problem of the full-order time-delay SPS is formulated. The singular perturbation techniques are used to obtain the reduced-order subsystems and equivalent subproblems. Then, a state reconstruction method is presented to represent the state in terms of available input and output data. In addition, a reinforcement learning (RL) algorithm based on input and output data is developed. Theoretical analyses are discussed on the convergence of the proposed algorithm and the asymptotical stability of the system under the composite controller. Finally, numerical simulation results illustrate the effectiveness of the proposed approach.

Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://hdl.handle.net/10.1080/00207721.2025.2454408 (text/html)
Access to full text is restricted to subscribers.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:taf:tsysxx:v:56:y:2025:i:11:p:2664-2673

Ordering information: This journal article can be ordered from
http://www.tandfonline.com/pricing/journal/TSYS20

DOI: 10.1080/00207721.2025.2454408

Access Statistics for this article

International Journal of Systems Science is currently edited by Visakan Kadirkamanathan

More articles in International Journal of Systems Science from Taylor & Francis Journals
Bibliographic data for series maintained by Chris Longhurst ().