Galvanin, F; Quaglio, M; Fraga, E; (2017) Model-based design of experiments under structural model uncertainty. In: Espuña, A and Graells, M and Puigjaner, L, (eds.) Computer Aided Engineering. (pp. pp. 145-150). Elsevier: Barcelona, Spain.

Text ESCAPE27_MBDoE_for_PP_revised_MQ(2).pdf - Accepted Version Access restricted to UCL open access staff Download (458kB)

Abstract

The development of comprehensive models for describing kinetic phenomena typically requires the extensive employment of resources for both determining the model structure and to precisely estimate the set of model parameters. For this reason, approximated model structures are normally employed instead of the comprehensive ones. The estimation of parameters in an incorrect model is not a trivial task. The expected information for the estimation of the model parameters is evaluated through the Fisher Information Matrix (FIM). Model-based design of experiments (MBDoE) methods for parameter precision do not take into account the structural model uncertainty in the formulation of FIM-based information metrics. In this work, an Extended FIM (E-FIM) is proposed, which carries information about the incorrect model structure. Its properties are discussed through its application on a simulated case with a biomass growth model.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as CSVJSONXML

Downloads by country - last 12 months

Export as JSONXMLCSV

Archive Staff Only

View Item