Design of Experiments for Reactivity Ratio Estimation in Multicomponent Polymerizations Using the Error-In-Variables Approach

• Published in: Macromolecular theory and simulations Vol. 22; no. 5; pp. 261 - 272
• Main Authors: Kazemi, Niousha, Duever, Thomas A., Penlidis, Alexander
• Format: Journal Article
• Language: English
• Published: Blackwell Publishing Ltd 01.06.2013
• Subjects: Copolymerization; Criteria; Design engineering; Design of experiments; Estimating; EVM parameter estimation; Mathematical models; model-based design of experiments; modeling; Optimization; reactivity ratios; Regression
• ISSN: 1022-1344; 1521-3919
• DOI: 10.1002/mats.201200085

Model‐based design of experiments using the error‐in‐variables model (EVM) is explored. The fundamental differences between DOE in the traditional nonlinear regression versus the EVM context are discussed, and it is pointed out that for cases where there are errors in all variables, using the EVM design criterion is the only appropriate approach. In addition, the implementation of the EVM design criterion and its characteristics for both initial and sequential design schemes are discussed. The main application is the implementation of the EVM criterion to design optimal trials for reliable estimating reactivity ratios for typical copolymerization systems, along with prescriptions for the practitioner. A detailed investigation of model‐based DOE under the EVM context shows that for cases where there are errors in all variables the EVM design criterion is the only appropriate approach. The implementation of the EVM design criterion is discussed, along with its application for designing optimal trials for estimating reactivity ratios for a typical copolymerization.