Sensitivity analysis for trajectories of nonsmooth mechanical systems with simultaneous impacts: A hybrid systems perspective

Mark Rijnen, Hao Liang Chen, Nathan Van De Wouw, Alessandro Saccon, Henk Nijmeijer

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review

8 Citations (Scopus)

Abstract

Sensitivity analysis for hybrid systems with state-triggered jumps is experiencing renewed attention for the control of robots with intermittent contacts. The basic assumption that enables this type of analysis is that jumps are triggered when the state reaches, transversally, a sufficiently smooth switching surface. In many scenarios of practical relevance, however, this switching surface is just piecewise smooth and, moreover, a perturbation of the initial conditions or the input leads to a different number of jumps than the nominal trajectory's. This work extends the sensitivity analysis in this context, under the assumptions that (i) at least locally, the intermediate perturbation-dependent jumps lead the system to reach always the nominal post-impact mode and (ii) once a switching and corresponding intermediate jump has occurred, its corresponding constraint remains active until reaching the nominal post-impact mode. Numerical simulations complement and validate the theoretical findings.

Original languageEnglish
Title of host publicationProceedings of the 2019 American Control Conference (ACC 2019)
Place of PublicationPiscataway, NJ, USA
PublisherIEEE
Pages3623-3629
ISBN (Electronic)978-1-5386-7926-5
DOIs
Publication statusPublished - 2019
Event2019 American Control Conference, ACC 2019 - Philadelphia, United States
Duration: 10 Jul 201912 Jul 2019

Conference

Conference2019 American Control Conference, ACC 2019
Country/TerritoryUnited States
CityPhiladelphia
Period10/07/1912/07/19

Fingerprint

Dive into the research topics of 'Sensitivity analysis for trajectories of nonsmooth mechanical systems with simultaneous impacts: A hybrid systems perspective'. Together they form a unique fingerprint.

Cite this