6533b834fe1ef96bd129d512

RESEARCH PRODUCT

Optimization of two-step batch processes and the method of compensation for random error

Tormod NæsKirsten BjørkestølEdvard Sivertsen

subject

Mathematical optimizationComputer scienceApplied Mathematicsmedia_common.quotation_subjectTwo stepEmphasis (telecommunications)Process (computing)Control variableAnalytical ChemistryCompensation (engineering)Set (abstract data type)Random errorQuality (business)media_common

description

This paper considers the problem of the optimal setting of controllable variables in two-step processes with quality constraints. The optimal setting minimizes the cost and satisfies quality constraints defined for the final output. The main emphasis is given to processes where it is possible to make intermediate measurements after the first processing step and to utilize these measurements before the control variables in the second step are set. Optimization based on this method of compensation for random error can yield substantially lower cost than does optimization based on a strategy where all variables are fixed before the process starts. An example of application of the method is taken from the food industry. Copyright © 2012 John Wiley & Sons, Ltd.

yearjournalcountryeditionlanguage
2012-05-03Journal of Chemometrics
https://doi.org/10.1002/cem.2449