Sparse solutions in optimal control of PDEs with uncertain parameters

The linear case

Chen Li, Georg Stadler

Research output: Contribution to journalArticle

Abstract

We study sparse solutions of optimal control problems governed by PDEs with uncertain coefficients. We propose two formulations, one where the solution is a deterministic control optimizing the mean objective, and a formulation aiming at stochastic controls that share the same sparsity structure. In both formulations, regions where the controls do not vanish can be interpreted as optimal locations for placing control devices. In this paper, we focus on linear PDEs with linearly entering uncertain parameters. Under these assumptions, the deterministic formulation reduces to a problem with known structure, and thus we mainly focus on the stochastic control formulation. Here, shared sparsity is achieved by incorporating the L 1 -norm of the mean of the pointwise squared controls in the objective. We reformulate the problem using a norm reweighting function that is defined over physical space only and thus helps to avoid approximation of the random space using samples or quadrature. We show that a fixed point algorithm applied to the norm reweighting formulation leads to a variant of the well-studied iterative reweighted least squares (IRLS) algorithm, and we propose a novel preconditioned Newton-conjugate gradient method to speed up the IRLS algorithm. We combine our algorithms with low-rank operator approximations, for which we provide estimates of the truncation error. We carefully examine the computational complexity of the resulting algorithms. The sparsity structure of the optimal controls and the performance of the solution algorithms are studied numerically using control problems governed by the Laplace and Helmholtz equations. In these experiments the Newton variant clearly outperforms the IRLS method.

Original languageEnglish (US)
Pages (from-to)633-658
Number of pages26
JournalSIAM Journal on Control and Optimization
Volume57
Issue number1
DOIs
StatePublished - Jan 1 2019

Fingerprint

Uncertain Parameters
Optimal Control
Formulation
Sparsity
Least Square Algorithm
Stochastic Control
Norm
Fixed-point Algorithm
Approximation Operators
Sample space
Optimal Location
Truncation Error
Conjugate Gradient Method
Helmholtz Equation
Laplace's equation
Least Square Method
Quadrature
Optimal Control Problem
Vanish
Control Problem

Keywords

  • Iterative reweighting
  • L minimization
  • Newton method
  • Optimal control of PDEs
  • Sparse controls
  • Uncertainty

ASJC Scopus subject areas

  • Control and Optimization
  • Applied Mathematics

Cite this

Sparse solutions in optimal control of PDEs with uncertain parameters : The linear case. / Li, Chen; Stadler, Georg.

In: SIAM Journal on Control and Optimization, Vol. 57, No. 1, 01.01.2019, p. 633-658.

Research output: Contribution to journalArticle

@article{ebe5e34c1d474b7b8524e78516d63fa7,
title = "Sparse solutions in optimal control of PDEs with uncertain parameters: The linear case",
abstract = "We study sparse solutions of optimal control problems governed by PDEs with uncertain coefficients. We propose two formulations, one where the solution is a deterministic control optimizing the mean objective, and a formulation aiming at stochastic controls that share the same sparsity structure. In both formulations, regions where the controls do not vanish can be interpreted as optimal locations for placing control devices. In this paper, we focus on linear PDEs with linearly entering uncertain parameters. Under these assumptions, the deterministic formulation reduces to a problem with known structure, and thus we mainly focus on the stochastic control formulation. Here, shared sparsity is achieved by incorporating the L 1 -norm of the mean of the pointwise squared controls in the objective. We reformulate the problem using a norm reweighting function that is defined over physical space only and thus helps to avoid approximation of the random space using samples or quadrature. We show that a fixed point algorithm applied to the norm reweighting formulation leads to a variant of the well-studied iterative reweighted least squares (IRLS) algorithm, and we propose a novel preconditioned Newton-conjugate gradient method to speed up the IRLS algorithm. We combine our algorithms with low-rank operator approximations, for which we provide estimates of the truncation error. We carefully examine the computational complexity of the resulting algorithms. The sparsity structure of the optimal controls and the performance of the solution algorithms are studied numerically using control problems governed by the Laplace and Helmholtz equations. In these experiments the Newton variant clearly outperforms the IRLS method.",
keywords = "Iterative reweighting, L minimization, Newton method, Optimal control of PDEs, Sparse controls, Uncertainty",
author = "Chen Li and Georg Stadler",
year = "2019",
month = "1",
day = "1",
doi = "10.1137/18M1181419",
language = "English (US)",
volume = "57",
pages = "633--658",
journal = "SIAM Journal on Control and Optimization",
issn = "0363-0129",
publisher = "Society for Industrial and Applied Mathematics Publications",
number = "1",

}

TY - JOUR

T1 - Sparse solutions in optimal control of PDEs with uncertain parameters

T2 - The linear case

AU - Li, Chen

AU - Stadler, Georg

PY - 2019/1/1

Y1 - 2019/1/1

N2 - We study sparse solutions of optimal control problems governed by PDEs with uncertain coefficients. We propose two formulations, one where the solution is a deterministic control optimizing the mean objective, and a formulation aiming at stochastic controls that share the same sparsity structure. In both formulations, regions where the controls do not vanish can be interpreted as optimal locations for placing control devices. In this paper, we focus on linear PDEs with linearly entering uncertain parameters. Under these assumptions, the deterministic formulation reduces to a problem with known structure, and thus we mainly focus on the stochastic control formulation. Here, shared sparsity is achieved by incorporating the L 1 -norm of the mean of the pointwise squared controls in the objective. We reformulate the problem using a norm reweighting function that is defined over physical space only and thus helps to avoid approximation of the random space using samples or quadrature. We show that a fixed point algorithm applied to the norm reweighting formulation leads to a variant of the well-studied iterative reweighted least squares (IRLS) algorithm, and we propose a novel preconditioned Newton-conjugate gradient method to speed up the IRLS algorithm. We combine our algorithms with low-rank operator approximations, for which we provide estimates of the truncation error. We carefully examine the computational complexity of the resulting algorithms. The sparsity structure of the optimal controls and the performance of the solution algorithms are studied numerically using control problems governed by the Laplace and Helmholtz equations. In these experiments the Newton variant clearly outperforms the IRLS method.

AB - We study sparse solutions of optimal control problems governed by PDEs with uncertain coefficients. We propose two formulations, one where the solution is a deterministic control optimizing the mean objective, and a formulation aiming at stochastic controls that share the same sparsity structure. In both formulations, regions where the controls do not vanish can be interpreted as optimal locations for placing control devices. In this paper, we focus on linear PDEs with linearly entering uncertain parameters. Under these assumptions, the deterministic formulation reduces to a problem with known structure, and thus we mainly focus on the stochastic control formulation. Here, shared sparsity is achieved by incorporating the L 1 -norm of the mean of the pointwise squared controls in the objective. We reformulate the problem using a norm reweighting function that is defined over physical space only and thus helps to avoid approximation of the random space using samples or quadrature. We show that a fixed point algorithm applied to the norm reweighting formulation leads to a variant of the well-studied iterative reweighted least squares (IRLS) algorithm, and we propose a novel preconditioned Newton-conjugate gradient method to speed up the IRLS algorithm. We combine our algorithms with low-rank operator approximations, for which we provide estimates of the truncation error. We carefully examine the computational complexity of the resulting algorithms. The sparsity structure of the optimal controls and the performance of the solution algorithms are studied numerically using control problems governed by the Laplace and Helmholtz equations. In these experiments the Newton variant clearly outperforms the IRLS method.

KW - Iterative reweighting

KW - L minimization

KW - Newton method

KW - Optimal control of PDEs

KW - Sparse controls

KW - Uncertainty

UR - http://www.scopus.com/inward/record.url?scp=85062714494&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85062714494&partnerID=8YFLogxK

U2 - 10.1137/18M1181419

DO - 10.1137/18M1181419

M3 - Article

VL - 57

SP - 633

EP - 658

JO - SIAM Journal on Control and Optimization

JF - SIAM Journal on Control and Optimization

SN - 0363-0129

IS - 1

ER -