Data-driven MPC applied to non-linear systems for real-time applications - Archive ouverte HAL Access content directly
Conference Papers Year : 2023

Data-driven MPC applied to non-linear systems for real-time applications

Abstract

Model Predictive Control (MPC) is a traditional technique widely employed on the control of constrained non-linear systems. In light of the increasing popularity of neural networks, data-driven MPC has emerged as an alternative to alleviate the computation burden of traditional control strategies. This paper aims to replace the constrained optimization problem by training a feed-forward neural network with data collected from an offline MPC simulation. The network's performance is then compared to that of traditional MPC using the Van der Pol oscillator as toy example. Finally, the convergence of the network training error is analytically proven by extending the analysis of neural tangent kernels (NTK) to underparameterized networks.
Fichier principal
Vignette du fichier
MSR_2024_Daniel_MARTIN.pdf (600.86 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-04423239 , version 1 (29-01-2024)

Identifiers

  • HAL Id : hal-04423239 , version 1

Cite

Daniel Martin Xavier, Ludovic Chamoin, Laurent Fribourg. Data-driven MPC applied to non-linear systems for real-time applications. Modélisation des Systèmes Réactifs (MSR'23), Nov 2023, Toulouse, France. ⟨hal-04423239⟩
8 View
25 Download

Share

Gmail Mastodon Facebook X LinkedIn More