Gaussian process-based nonlinearity compensation for pneumatic soft actuators

Research output: Contribution to journalArticleResearchpeer review

Authors

  • Alexander Pawluchin
  • Michael Meindl
  • Ive Weygers
  • Thomas Seel
  • Ivo Boblan

Research Organisations

External Research Organisations

  • Berliner Hochschule für Technik (BHT)
  • Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU Erlangen-Nürnberg)
View graph of relations

Details

Translated title of the contributionEine auf Gauß-Prozessen basierende Kompensation der Nichtlinearitäten für weiche pneumatische Aktuatoren
Original languageEnglish
Pages (from-to)440-448
Number of pages9
JournalAt-Automatisierungstechnik
Volume72
Issue number5
Early online date7 May 2024
Publication statusPublished - 27 May 2024

Abstract

Highly compliant Pneumatic Soft Actuators (PSAs) have the potential to perform challenging tasks in a broad range of applications that require shape-adaptive capabilities. Achieving accurate tracking control for such actuators with complex geometries and material compositions typically involves many time-consuming and laborious engineering steps. In this work, we propose a data-driven learning-based control approach to address reference tracking tasks, incorporating self-adaptation in situ. We utilize a short interaction maneuver, recorded a priori, to collect the quasi-static data affected by severe hysteresis. Besides a linear feedback controller, we use two Gaussian process models to predict the feedforward control input to compensate for the nonlinearity in a one-shot learning setting. The proposed control approach demonstrates accurate tracking performance even under realistic varying configurations, such as alterations in mass and orientation, without any parameter tuning. Notably, training was achieved with only 25-50 s of experimental interaction, which emphasizes the plug-and-play capabilities in diverse real-world applications.

Keywords

    feedforward control, Gaussian process, hysteresis modeling, pneumatic soft actuator, reference tracking, soft robotics

ASJC Scopus subject areas

Cite this

Gaussian process-based nonlinearity compensation for pneumatic soft actuators. / Pawluchin, Alexander; Meindl, Michael; Weygers, Ive et al.
In: At-Automatisierungstechnik, Vol. 72, No. 5, 27.05.2024, p. 440-448.

Research output: Contribution to journalArticleResearchpeer review

Pawluchin, A, Meindl, M, Weygers, I, Seel, T & Boblan, I 2024, 'Gaussian process-based nonlinearity compensation for pneumatic soft actuators', At-Automatisierungstechnik, vol. 72, no. 5, pp. 440-448. https://doi.org/10.1515/auto-2023-0237
Pawluchin A, Meindl M, Weygers I, Seel T, Boblan I. Gaussian process-based nonlinearity compensation for pneumatic soft actuators. At-Automatisierungstechnik. 2024 May 27;72(5):440-448. Epub 2024 May 7. doi: 10.1515/auto-2023-0237
Pawluchin, Alexander ; Meindl, Michael ; Weygers, Ive et al. / Gaussian process-based nonlinearity compensation for pneumatic soft actuators. In: At-Automatisierungstechnik. 2024 ; Vol. 72, No. 5. pp. 440-448.
Download
@article{8299676579dc41c3ba42e80a7b5c7de4,
title = "Gaussian process-based nonlinearity compensation for pneumatic soft actuators",
abstract = "Highly compliant Pneumatic Soft Actuators (PSAs) have the potential to perform challenging tasks in a broad range of applications that require shape-adaptive capabilities. Achieving accurate tracking control for such actuators with complex geometries and material compositions typically involves many time-consuming and laborious engineering steps. In this work, we propose a data-driven learning-based control approach to address reference tracking tasks, incorporating self-adaptation in situ. We utilize a short interaction maneuver, recorded a priori, to collect the quasi-static data affected by severe hysteresis. Besides a linear feedback controller, we use two Gaussian process models to predict the feedforward control input to compensate for the nonlinearity in a one-shot learning setting. The proposed control approach demonstrates accurate tracking performance even under realistic varying configurations, such as alterations in mass and orientation, without any parameter tuning. Notably, training was achieved with only 25-50 s of experimental interaction, which emphasizes the plug-and-play capabilities in diverse real-world applications.",
keywords = "feedforward control, Gaussian process, hysteresis modeling, pneumatic soft actuator, reference tracking, soft robotics",
author = "Alexander Pawluchin and Michael Meindl and Ive Weygers and Thomas Seel and Ivo Boblan",
note = "Publisher Copyright: {\textcopyright} 2024 Walter de Gruyter GmbH, Berlin/Boston.",
year = "2024",
month = may,
day = "27",
doi = "10.1515/auto-2023-0237",
language = "English",
volume = "72",
pages = "440--448",
journal = "At-Automatisierungstechnik",
issn = "0178-2312",
publisher = "Walter de Gruyter GmbH",
number = "5",

}

Download

TY - JOUR

T1 - Gaussian process-based nonlinearity compensation for pneumatic soft actuators

AU - Pawluchin, Alexander

AU - Meindl, Michael

AU - Weygers, Ive

AU - Seel, Thomas

AU - Boblan, Ivo

N1 - Publisher Copyright: © 2024 Walter de Gruyter GmbH, Berlin/Boston.

PY - 2024/5/27

Y1 - 2024/5/27

N2 - Highly compliant Pneumatic Soft Actuators (PSAs) have the potential to perform challenging tasks in a broad range of applications that require shape-adaptive capabilities. Achieving accurate tracking control for such actuators with complex geometries and material compositions typically involves many time-consuming and laborious engineering steps. In this work, we propose a data-driven learning-based control approach to address reference tracking tasks, incorporating self-adaptation in situ. We utilize a short interaction maneuver, recorded a priori, to collect the quasi-static data affected by severe hysteresis. Besides a linear feedback controller, we use two Gaussian process models to predict the feedforward control input to compensate for the nonlinearity in a one-shot learning setting. The proposed control approach demonstrates accurate tracking performance even under realistic varying configurations, such as alterations in mass and orientation, without any parameter tuning. Notably, training was achieved with only 25-50 s of experimental interaction, which emphasizes the plug-and-play capabilities in diverse real-world applications.

AB - Highly compliant Pneumatic Soft Actuators (PSAs) have the potential to perform challenging tasks in a broad range of applications that require shape-adaptive capabilities. Achieving accurate tracking control for such actuators with complex geometries and material compositions typically involves many time-consuming and laborious engineering steps. In this work, we propose a data-driven learning-based control approach to address reference tracking tasks, incorporating self-adaptation in situ. We utilize a short interaction maneuver, recorded a priori, to collect the quasi-static data affected by severe hysteresis. Besides a linear feedback controller, we use two Gaussian process models to predict the feedforward control input to compensate for the nonlinearity in a one-shot learning setting. The proposed control approach demonstrates accurate tracking performance even under realistic varying configurations, such as alterations in mass and orientation, without any parameter tuning. Notably, training was achieved with only 25-50 s of experimental interaction, which emphasizes the plug-and-play capabilities in diverse real-world applications.

KW - feedforward control

KW - Gaussian process

KW - hysteresis modeling

KW - pneumatic soft actuator

KW - reference tracking

KW - soft robotics

UR - http://www.scopus.com/inward/record.url?scp=85192832543&partnerID=8YFLogxK

U2 - 10.1515/auto-2023-0237

DO - 10.1515/auto-2023-0237

M3 - Article

AN - SCOPUS:85192832543

VL - 72

SP - 440

EP - 448

JO - At-Automatisierungstechnik

JF - At-Automatisierungstechnik

SN - 0178-2312

IS - 5

ER -