Task-Driven Hybrid Model Reduction for Dexterous Manipulation
IEEE Transactions on Robotics · 2024 · 40 · 1774-1794
DOI: 10.1109/tro.2024.3359531 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Robotics, 2024. Verfasst von Wanxin Jin und 1 weiteren Autor:innen. Die Arbeit wurde bisher 13-mal zitiert (Crossref) und nennt 49 Literaturangaben. Gefördert von: Toyota Research Institute.
Abstract (Originalsprache)
In contact-rich tasks, like dexterous manipulation, the hybrid nature of making and breaking contact creates challenges for model representation and control. For example, choosing and sequencing contact locations for in-hand manipulation, where there are thousands of potential hybrid modes, is not generally tractable. In this article, we are inspired by the observation that far fewer modes are actually necessary to accomplish many tasks. Building on our prior work learning hybrid models, represented as linear complementarity systems, we find a reduced-order hybrid model requiring only a limited number of task-relevant modes. This simplified representation, in combination with model predictive control, enables real-time control yet is sufficient for achieving high performance. We demonstrate the proposed method first on synthetic hybrid systems, reducing the mode count by multiple orders of magnitude while achieving task performance loss of less than 5%. We also apply the proposed method to a three-fingered robotic hand manipulating a previously unknown object. With no prior knowledge, we achieve state-of-the-art closed-loop performance within a few minutes of online learning, by collecting only a few thousand environment samples.
Förderung
- Toyota Research Institute
Zitieren
APA
Jin, W., & Posa, M. (2024). Task-Driven Hybrid Model Reduction for Dexterous Manipulation. IEEE Transactions on Robotics, 40, 1774-1794. https://doi.org/10.1109/tro.2024.3359531
Vancouver
Jin W, Posa M. Task-Driven Hybrid Model Reduction for Dexterous Manipulation. IEEE Transactions on Robotics. 2024;40:1774-1794. doi:10.1109/tro.2024.3359531
BibTeX
@article{Jin2024,
author = {Jin, Wanxin and Posa, Michael},
title = {Task-Driven Hybrid Model Reduction for Dexterous Manipulation},
journal = {IEEE Transactions on Robotics},
year = {2024},
volume = {40},
pages = {1774--1794},
doi = {10.1109/tro.2024.3359531}
}
EUVIMED