Ensemble-marginalized Kalman filter for linear time-dependent PDEs with noisy boundary conditions: Application to heat transfer in building walls

Handle URI:
http://hdl.handle.net/10754/626490
Title:
Ensemble-marginalized Kalman filter for linear time-dependent PDEs with noisy boundary conditions: Application to heat transfer in building walls
Authors:
Iglesias, Marco; Sawlan, Zaid A ( 0000-0002-4361-2491 ) ; Scavino, Marco ( 0000-0001-5114-853X ) ; Tempone, Raul ( 0000-0003-1967-4446 ) ; Wood, Christopher
Abstract:
In this work, we present the ensemble-marginalized Kalman filter (EnMKF), a sequential algorithm analogous to our previously proposed approach [1,2], for estimating the state and parameters of linear parabolic partial differential equations in initial-boundary value problems when the boundary data are noisy. We apply EnMKF to infer the thermal properties of building walls and to estimate the corresponding heat flux from real and synthetic data. Compared with a modified Ensemble Kalman Filter (EnKF) that is not marginalized, EnMKF reduces the bias error, avoids the collapse of the ensemble without needing to add inflation, and converges to the mean field posterior using $50\%$ or less of the ensemble size required by EnKF. According to our results, the marginalization technique in EnMKF is key to performance improvement with smaller ensembles at any fixed time.
KAUST Department:
Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
Publisher:
arXiv
KAUST Grant Number:
2281; 2584
Issue Date:
26-Nov-2017
ARXIV:
arXiv:1711.09365
Type:
Preprint
Additional Links:
http://arxiv.org/abs/1711.09365v1; http://arxiv.org/pdf/1711.09365v1
Appears in Collections:
Other/General Submission; Other/General Submission; Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division

Full metadata record

DC FieldValue Language
dc.contributor.authorIglesias, Marcoen
dc.contributor.authorSawlan, Zaid Aen
dc.contributor.authorScavino, Marcoen
dc.contributor.authorTempone, Raulen
dc.contributor.authorWood, Christopheren
dc.date.accessioned2017-12-28T07:32:13Z-
dc.date.available2017-12-28T07:32:13Z-
dc.date.issued2017-11-26en
dc.identifier.urihttp://hdl.handle.net/10754/626490-
dc.description.abstractIn this work, we present the ensemble-marginalized Kalman filter (EnMKF), a sequential algorithm analogous to our previously proposed approach [1,2], for estimating the state and parameters of linear parabolic partial differential equations in initial-boundary value problems when the boundary data are noisy. We apply EnMKF to infer the thermal properties of building walls and to estimate the corresponding heat flux from real and synthetic data. Compared with a modified Ensemble Kalman Filter (EnKF) that is not marginalized, EnMKF reduces the bias error, avoids the collapse of the ensemble without needing to add inflation, and converges to the mean field posterior using $50\%$ or less of the ensemble size required by EnKF. According to our results, the marginalization technique in EnMKF is key to performance improvement with smaller ensembles at any fixed time.en
dc.publisherarXiven
dc.relation.urlhttp://arxiv.org/abs/1711.09365v1en
dc.relation.urlhttp://arxiv.org/pdf/1711.09365v1en
dc.rightsArchived with thanks to arXiven
dc.titleEnsemble-marginalized Kalman filter for linear time-dependent PDEs with noisy boundary conditions: Application to heat transfer in building wallsen
dc.typePreprinten
dc.contributor.departmentComputer, Electrical and Mathematical Sciences and Engineering (CEMSE) Divisionen
dc.eprint.versionPre-printen
dc.contributor.institutionSchool of Mathematical Sciences, University of Nottingham, Nottingham, UKen
dc.contributor.institutionInstituto de Estad ́ıstica (IESTA), Universidad de la Rep ́ublica, Montevideo, Uruguen
dc.contributor.institutionDepartment of Architecture and Built Environment, University of Nottingham, Nottingham, UKen
dc.identifier.arxividarXiv:1711.09365en
kaust.authorSawlan, Zaid Aen
kaust.authorScavino, Marcoen
kaust.authorTempone, Raulen
kaust.grant.number2281en
kaust.grant.number2584en
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.