Skip to main navigation Skip to search Skip to main content

Data Driven Sliding Mode Control: A Model-Free approach

  • Hanna Aboukheir*
  • *Corresponding author for this work

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

3 Scopus citations

Abstract

Sliding mode control (SMC) is a robust control strategy that is effective against disturbances, delays and model uncertainties; but similar to other control strategies, SMC is a model-based control method, and the complexity of certain industrial processes makes model estimation a highly difficult task. Furthermore, the growing interest on Industry 4.0 and the new requirements for control loops raise some questions about the future role of this controller in the industrial field. In this work a design methodology for sliding mode control based on data-driven control concepts is presented. Beginning with an extended-state Kalman filter for unknown output estimation, with this information the discrete-time reachability law is derived for the calculation of the control law. Finally, the proposal is evaluated in a simulated mixing plant and implemented on the TCLAB process.

Original languageEnglish
Title of host publicationECTM 2023 - 2023 IEEE 7th Ecuador Technical Chapters Meeting
EditorsDavid Rivas Lalaleo, Manuel Ignacio Ayala Chauvin
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798350338232
DOIs
StatePublished - 2023
Event7th IEEE Ecuador Technical Chapters Meeting, ECTM 2023 - Ambato, Ecuador
Duration: 10 Oct 202313 Oct 2023

Publication series

NameECTM 2023 - 2023 IEEE 7th Ecuador Technical Chapters Meeting

Conference

Conference7th IEEE Ecuador Technical Chapters Meeting, ECTM 2023
Country/TerritoryEcuador
CityAmbato
Period10/10/2313/10/23

Keywords

  • Kalman Filter
  • Sliding mode control
  • State estimation
  • data-driven systems
  • model-free

Fingerprint

Dive into the research topics of 'Data Driven Sliding Mode Control: A Model-Free approach'. Together they form a unique fingerprint.

Cite this