RUC. Repositorio da Universidade da Coruña
Publication
Artículo científico (JournalArticle)
Adjoint method for parameter identification problems in models of stirred tank chemical reactors
[Abstract]: In this paper an adjoint-based algorithm for parameter identification problems in systems of ordinary differential equations (ODEs) is presented. This is done by solving a minimization problem in which the cost function is defined in order to quantify the mismatch between the observed data and the numerical solution of the ODEs. Most of existing local minimizers need the derivatives of both the cost function and the constraints that are usually calculated by finite-difference formulas. In this paper we show how they can be computed much more efficiently and accurately by the so-called adjoint method. We apply this method to the problem of estimating a set of unknown parameters appearing in chemical reaction models. Numerical results showing the efficiency of the adjoint method are included., This research was developed as an activity in the Joint Research Unit Repsol-ITMATI (code file: IN853A 2014/03) which is funded by the Galician Agency for Innovation and the Ministry of Economy and Competitiveness in the framework of the Spanish Strategy for Innovation in Galicia. The first author has been also financed by FEDER and Spanish Ministry of Economy, Industry and Competitiveness under research project MTM2013-43745-R, and by FEDER and Xunta de Galicia, and the second author has been also financed by FEDER and Spanish Ministry of Science and Innovation under research project ENE2013-47867-C2-1-R and by FEDER and Xunta de Galicia under research project GRC2013/014., Axencia Galega de Innovación; IN853A 2014/03, Xunta de Galicia; GRC2013/014
Project:
MINECO / Plan Estatal de Investigación Científica y Técnica y de Innovación 2013-2016 MTM2013-43745-R
,
MINECO / Plan Estatal de Investigación Científica y Técnica y de Innovación 2013-2016 ENE2013-47867-C2-1-R
HANDLE: http://hdl.handle.net/2183/40800
View at: http://hdl.handle.net/2183/40800