Numerical issues and turnpike phenomenon in optimal shape design - Inria - Institut national de recherche en sciences et technologies du numérique
Book Sections Year : 2022

Numerical issues and turnpike phenomenon in optimal shape design

Abstract

This article follows and complements where we have established the turnpike property for some optimal shape design problems. Considering linear parabolic partial differential equations where the shapes to be optimized acts as a source term, we want to minimize a quadratic criterion. Existence of optimal shapes is proved under some appropriate assumptions. We prove and provide numerical evidence of the turnpike phenomenon for those optimal shapes, meaning that the extremal time-varying optimal solution remains essentially stationary; actually, it remains essentially close to the optimal solution of an associated static problem.
Fichier principal
Vignette du fichier
LTZ_RICAM.pdf (3.78 Mo) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-03660786 , version 1 (06-05-2022)

Identifiers

Cite

Gontran Lance, Emmanuel Trélat, Enrique Zuazua. Numerical issues and turnpike phenomenon in optimal shape design. Roland Herzog; Matthias Heinkenschloss; Dante Kalise; Georg Stadler; Emmanuel Trélat. Optimization and control for partial differential equations: Uncertainty quantification, open and closed-loop control, and shape optimization, 29, De Gruyter, pp.343-366, 2022, Radon Series on Computational and Applied Mathematics, 9783110695960. ⟨10.1515/9783110695984⟩. ⟨hal-03660786⟩
76 View
88 Download

Altmetric

Share

More