Physics-informed Gaussian Processes as Linear Model Predictive Controller

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Tebbe, Jörn, Besginow, Andreas, Lange-Hegermann, Markus
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909713285775360
author Tebbe, Jörn
Besginow, Andreas
Lange-Hegermann, Markus
author_facet Tebbe, Jörn
Besginow, Andreas
Lange-Hegermann, Markus
contents We introduce a novel algorithm for controlling linear time invariant systems in a tracking problem. The controller is based on a Gaussian Process (GP) whose realizations satisfy a system of linear ordinary differential equations with constant coefficients. Control inputs for tracking are determined by conditioning the prior GP on the setpoints, i.e. control as inference. The resulting Model Predictive Control scheme incorporates pointwise soft constraints by introducing virtual setpoints to the posterior Gaussian process. We show theoretically that our controller satisfies open-loop stability for the optimal control problem by leveraging general results from Bayesian inference and demonstrate this result in a numerical example.
format Preprint
id arxiv_https___arxiv_org_abs_2412_04502
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Physics-informed Gaussian Processes as Linear Model Predictive Controller
Tebbe, Jörn
Besginow, Andreas
Lange-Hegermann, Markus
Optimization and Control
Machine Learning
Systems and Control
We introduce a novel algorithm for controlling linear time invariant systems in a tracking problem. The controller is based on a Gaussian Process (GP) whose realizations satisfy a system of linear ordinary differential equations with constant coefficients. Control inputs for tracking are determined by conditioning the prior GP on the setpoints, i.e. control as inference. The resulting Model Predictive Control scheme incorporates pointwise soft constraints by introducing virtual setpoints to the posterior Gaussian process. We show theoretically that our controller satisfies open-loop stability for the optimal control problem by leveraging general results from Bayesian inference and demonstrate this result in a numerical example.
title Physics-informed Gaussian Processes as Linear Model Predictive Controller
topic Optimization and Control
Machine Learning
Systems and Control
url https://arxiv.org/abs/2412.04502