Variable-impedance muscle coordination under slow-rate control frequencies and limited observation conditions evaluated through legged locomotion
Bioinspiration & Biomimetics · 31.08.2026 · 21 (4) · 046037
DOI: 10.1088/1748-3190/ae8a9e ↗Kurz gefasst
Fachartikel in der Zeitschrift Bioinspiration & Biomimetics, online erschienen am 19.08.2026. Verfasst von Hidaka Asai und 2 weiteren Autor:innen. Beteiligte Einrichtung: Brain Robot Interface, ATR Computational Neuroscience Labs, Kyoto, Japan. Hauptthemen laut MEDLINE: Muscle, Skeletal, Locomotion. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Tateishi Science and Technology Foundation, JAPAN, Japan Science and Technology Agency (JST) Moonshot R&D.
Abstract (Originalsprache)
Human motor control remains agile and robust despite limited sensory information for feedback, a property attributed to the body’s ability to perform morphological computation through muscle coordination (MC) with variable impedance. However, it remains unclear how such low-level mechanical computation reduces the requirements of the high-level controller. In this study, we implement a hierarchical controller consisting of a high-level neural network trained by reinforcement learning and a low-level variable-impedance MC model with mono- and biarticular muscles in legged locomotion task. We systematically restrict the high-level controller by varying the control frequency and by introducing biologically inspired observation conditions: delayed, partial, and substituted observation. Under these conditions, we evaluate how the low-level variable-impedance MC contributes to learning process of high-level neural network. The results show that variable-impedance MC enables stable locomotion even under slow-rate control frequency and limited observation conditions. These findings demonstrate that the morphological computation of MC effectively offloads high-frequency feedback of the high-level controller and provide a design principle for the controller in motor control.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Brain Robot Interface, ATR Computational Neuroscience Labs, Kyoto, Japan.
- Learning Machines Group, Graduate School of Informatics, Kyoto University, Kyoto, Japan.
ORCID-Profile: Hidaka Asai, Tomoyuki Noda, Jun Morimoto
Förderung
- Tateishi Science and Technology Foundation
- JAPAN
- Japan Science and Technology Agency (JST) Moonshot R&D
- NEDO
- JSPS KAKENHI
Zitieren
APA
Asai, H., Noda, T., & Morimoto, J. (2026). Variable-impedance muscle coordination under slow-rate control frequencies and limited observation conditions evaluated through legged locomotion. Bioinspiration & Biomimetics, 21(4), 046037. https://doi.org/10.1088/1748-3190/ae8a9e
Vancouver
Asai H, Noda T, Morimoto J. Variable-impedance muscle coordination under slow-rate control frequencies and limited observation conditions evaluated through legged locomotion. Bioinspiration & Biomimetics. 2026;21(4):046037. doi:10.1088/1748-3190/ae8a9e
BibTeX
@article{Asai2026,
author = {Asai, Hidaka and Noda, Tomoyuki and Morimoto, Jun},
title = {Variable-impedance muscle coordination under slow-rate control frequencies and limited observation conditions evaluated through legged locomotion},
journal = {Bioinspiration & Biomimetics},
year = {2026},
volume = {21},
number = {4},
pages = {046037},
doi = {10.1088/1748-3190/ae8a9e}
}
EUVIMED