Premium
Experimental Design in Nonlinear Case Applied to Hydrocracking Model: How Many Points Do We Need and Which Ones?
Author(s) -
Celse Benoit,
Costa JeanJérôme Da,
Costa Victor
Publication year - 2016
Publication title -
international journal of chemical kinetics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.341
H-Index - 68
eISSN - 1097-4601
pISSN - 0538-8066
DOI - 10.1002/kin.21021
Subject(s) - selection (genetic algorithm) , experimental data , design of experiments , work (physics) , nonlinear system , data point , algorithm , computer science , mathematical optimization , statistics , mathematics , machine learning , mechanical engineering , engineering , physics , quantum mechanics
In this work, different experimental design techniques were applied to a hydrocracking conversion model (single response). The selection of points was carried out using Federov's algorithm from an experimental database containing 48 points. The sequential design methodology was also applied. Results show that the D ‐optimality criterion can provide valuable information for determination of the minimum number of experiments required to achieve acceptable precision. It is shown that only 22 points are enough to estimate parameters instead of the original database of 48 points. If sequential techniques are used, only 13 points are needed. This work provides useful theoretical and practical guidelines for experimental design in single‐response problems.