In the development of Model Predictive Control (MPC) systems, having a data-driven approach to automatically identify black-box dynamical models that are both accurate and computationally efficient is of paramount importance. In this work, we propose an extension of the learning algorithm proposed in [1] to reduce the state/model size during the training procedure by employing RESNETs and grouplasso, thus easing its usage in embedded MPC applications. Performance will be assessed on the Silverbox benchmark.

Nonlinear systems identification with automatic state/model size reduction using RESNETs and Group Lasso

L. Frascati;A. Bemporad
2023-01-01

Abstract

In the development of Model Predictive Control (MPC) systems, having a data-driven approach to automatically identify black-box dynamical models that are both accurate and computationally efficient is of paramount importance. In this work, we propose an extension of the learning algorithm proposed in [1] to reduce the state/model size during the training procedure by employing RESNETs and grouplasso, thus easing its usage in embedded MPC applications. Performance will be assessed on the Silverbox benchmark.
File in questo prodotto:
File Dimensione Formato  
Nonlinear systems identification with RESNETs.pdf

Accesso riservato

Descrizione: Abstract
: Altro materiale allegato
Dimensione 89.06 kB
Formato Adobe PDF
89.06 kB Adobe PDF   Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11311/1260365
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact