Learning Koopman Reduced-Order Model for Feedback Control in Dexterous Manipulation
IEEE Access · 2026 · 14 · 70912-70924
DOI: 10.1109/access.2026.3691534 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Mohamed MohamedAhmed und 3 weiteren Autor:innen. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: King Fahd University of Petroleum and Minerals (KFUPM)-Deanship of Research Oversight and Coordination, Interdisciplinary Research Center (IRC) for Smart Mobility and Logistics, KFUPM.
Abstract (Originalsprache)
We propose a data-driven feedback control method for dexterous manipulation that uses a Koopman-based reduced-order model (ROM) and a simple static state-feedback law in a lifted state space. From demonstrations, we learn Koopman reference dynamics, and from rollouts with control, we identify a controlled Koopman model using extended Dynamic Mode Decomposition with control (eDMDc); both are projected onto a common reduced subspace where the feedback gain is chosen to match the reduced closed-loop dynamics to the reduced reference. To ensure stability of the learned reference, we apply spectral clipping (SC) to enforce a desired discrete-time stability margin. We also derive a tracking-error bound that separates the effects of model mismatch, projection and invariance errors, and residual disturbances. The method is evaluated on tool-use and door-opening tasks in the MuJoCo Adroit hand environment, using third-order polynomial observables to study success rate, RMS tracking error, final distance to the goal, and controllable modes as functions of the reduced order, with and without SC, and achieves high success with microsecond-level online computation time.
Förderung
- King Fahd University of Petroleum and Minerals (KFUPM)-Deanship of Research Oversight and Coordination
- Interdisciplinary Research Center (IRC) for Smart Mobility and Logistics, KFUPM
Zitieren
APA
MohamedAhmed, M., Emzir, M. F., HASAN, A. I., & Rashad, R. (2026). Learning Koopman Reduced-Order Model for Feedback Control in Dexterous Manipulation. IEEE Access, 14, 70912-70924. https://doi.org/10.1109/access.2026.3691534
Vancouver
MohamedAhmed M, Emzir MF, HASAN AI, Rashad R. Learning Koopman Reduced-Order Model for Feedback Control in Dexterous Manipulation. IEEE Access. 2026;14:70912-70924. doi:10.1109/access.2026.3691534
BibTeX
@article{MohamedAhmed2026,
author = {MohamedAhmed, Mohamed and Emzir, Muhammad Fuady and HASAN, Agus Ismail and Rashad, Ramy},
title = {Learning Koopman Reduced-Order Model for Feedback Control in Dexterous Manipulation},
journal = {IEEE Access},
year = {2026},
volume = {14},
pages = {70912--70924},
doi = {10.1109/access.2026.3691534}
}
EUVIMED