Learning stable dynamical systems using contraction theory

Caroline Blocher, Matteo Saveriano, Dongheui Lee

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

28 Scopus citations

Abstract

This paper discusses the learning of robot point-to-point motions via non-linear dynamical systems and Gaussian Mixture Regression (GMR). The novelty of the proposed approach consists in guaranteeing the stability of a learned dynamical system via Contraction theory. A contraction analysis is performed to derive sufficient conditions for the global stability of a dynamical system represented by GMR. The results of this analysis are exploited to automatically compute a control input which stabilizes the learned system on-line. Simple and effective solutions are proposed to generate motion trajectories close to the demonstrated ones, without affecting the stability of the overall system. The proposed approach is evaluated on a public benchmark of point-to-point motions and compared with state-of-the-art algorithms based on Lyapunov stability theory.

Original languageEnglish
Title of host publication2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages124-129
Number of pages6
ISBN (Electronic)9781509030552
DOIs
StatePublished - 25 Jul 2017
Externally publishedYes
Event14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017 - Jeju, Korea, Republic of
Duration: 28 Jun 20171 Jul 2017

Publication series

Name2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017

Conference

Conference14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017
Country/TerritoryKorea, Republic of
CityJeju
Period28/06/171/07/17

Keywords

  • Learning contracting systems. Stable discrete movements. Learning from demonstration. Contraction theory

Fingerprint

Dive into the research topics of 'Learning stable dynamical systems using contraction theory'. Together they form a unique fingerprint.

Cite this