EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Discrete‐Time Caputo Fractional‐Order Terminal Sliding Mode Control for Uncertain MIMO Systems Using Multi‐Agent Reinforcement Learning With Disturbance Compensation

Amir Veisi, Hadi Delavari

Vector: The Journal of Engineering · 2027-01 · 2027 (1)

DOI: 10.1049/tje2.70238 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Vector: The Journal of Engineering, online erschienen am 09.11.2026. Verfasst von Amir Veisi und 1 weiteren Autor:innen. Die Arbeit nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This paper proposes a novel data‐driven discrete‐time control framework for uncertain nonlinear MIMO systems by integrating discrete‐time Caputo fractional‐order terminal sliding mode control, multi‐agent reinforcement learning and reinforcement learning‐based disturbance compensation. Unlike existing discrete‐time fractional‐order sliding mode controllers that rely on accurate system models and fixed controller parameters, the proposed framework simultaneously achieves adaptive controller optimisation and online disturbance compensation using only input‐output measurements, thereby eliminating the dependence on precise mathematical modelling. A discrete‐time Caputo fractional‐order terminal sliding surface is developed to guarantee finite‐time convergence while preserving the memory characteristics of fractional‐order dynamics. To improve robustness against external disturbances and internal uncertainties, a reinforcement learning‐based disturbance observer is designed to estimate and compensate for unknown perturbations in real time. Furthermore, a multi‐agent reinforcement learning strategy is employed to adaptively optimise the controller gains without manual parameter tuning, resulting in enhanced tracking performance under varying operating conditions. The closed‐loop stability of the proposed controller is rigorously established using Lyapunov theory for discrete‐time fractional‐order systems. Comparative simulation studies conducted on a benchmark nonlinear MIMO system demonstrate that the proposed framework achieves a 14.54% reduction in the integral of absolute error (IAE) under the model uncertainty‐only scenario compared with the conventional fractional‐order sliding mode controller. Furthermore, additional simulation results under combined model uncertainties and external disturbances confirm the superior robustness, adaptive capability and tracking accuracy of the proposed framework.

Zitieren

APA

Veisi, A., & Delavari, H. (2027). Discrete‐Time Caputo Fractional‐Order Terminal Sliding Mode Control for Uncertain MIMO Systems Using Multi‐Agent Reinforcement Learning With Disturbance Compensation. Vector: The Journal of Engineering, 2027(1). https://doi.org/10.1049/tje2.70238

Vancouver

Veisi A, Delavari H. Discrete‐Time Caputo Fractional‐Order Terminal Sliding Mode Control for Uncertain MIMO Systems Using Multi‐Agent Reinforcement Learning With Disturbance Compensation. Vector: The Journal of Engineering. 2027;2027(1). doi:10.1049/tje2.70238

BibTeX
@article{Veisi2027,
  author = {Veisi, Amir and Delavari, Hadi},
  title = {Discrete‐Time Caputo Fractional‐Order Terminal Sliding Mode Control for Uncertain MIMO Systems Using Multi‐Agent Reinforcement Learning With Disturbance Compensation},
  journal = {Vector: The Journal of Engineering},
  year = {2027},
  volume = {2027},
  number = {1},
  doi = {10.1049/tje2.70238}
}
Kostenlos starten →