Emulsion copolymerization of styrene and butyl acrylate in the presence of a chain transfer agent. Part 2: Parameters estimability and confidence regions

Accurate estimation of the model parameters is required to obtain reliable predictions of the products end-use properties. However, due to the mathematical model structure and/or to a possible lack of measurements, the estimation of some parameters may be impossible. This paper will focus on the cas...

Full description

Saved in:
Bibliographic Details
Published in:Chemical engineering science 2013-03, Vol.90 (7), p.110-118
Main Authors: Benyahia, B., Latifi, M.A., Fonteix, C., Pla, F.
Format: Article
Language:English
Subjects:
Acrylates
Chain transfer
chemical engineering
Chemical Sciences
Confidence
Confidence intervals
Copolymerization
Emulsion copolymerization
Emulsions
Estimability analysis
Mathematical models
Parameter ranking
Polymers
prediction
styrene
Styrenes
True confidence regions Robust estimability Genetic algorithm
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Accurate estimation of the model parameters is required to obtain reliable predictions of the products end-use properties. However, due to the mathematical model structure and/or to a possible lack of measurements, the estimation of some parameters may be impossible. This paper will focus on the case where the main limitations to the parameters estimability are their weak effect on the measured outputs or the correlation between the effects of two or more parameters. The objective of the method developed in this paper is to determine the subset of the most influencing parameters that can be estimated from the available experimental data, when the complete set of model parameters cannot be estimated. This approach has been applied to the mathematical model of the emulsion copolymerization of styrene and butyl acrylate, in the presence of n-dodecyl mercaptan as a chain transfer agent. In addition, a new approach is used to better assess the true confidence regions and evaluate the accuracy of the parameters estimates in more reliable way. ▸ Robust selection of the most influential and linearly independent parameters. ▸ An estimability analysis based on a sequential orthogonalization procedure. ▸ Selection of a set of 21 parameters among 49 parameters. ▸ Development of a novel approach to evaluate the true confidence regions. ▸ Combination of the estimability analysis and the true confidence regions.
ISSN:0009-2509
1873-4405
DOI:10.1016/j.ces.2012.12.013