Comparison of RSM and ANN for the investigation of linear alkylbenzene synthesis over Hl41[NaP5W30O110]/SiO2 catalyst

Design of experiments (DOE) and artificial neural networks (ANNs) were successfully applied for studying the operating parameters of benzene alkylation with 1-decene over H-14[NaP5W30O110]/SiO2 catalyst. In this reaction catalyst loading, catalyst weight percent and benzene to 1-decene molar ratio (Bz/C,(10)) were chosen as independent variables in experimental design. Prediction of 1-decene conversion and 2-phenyldecane selectivity was performed applying response surface method (RSM) and ANN models. Final selected multi-layer (3-6-2-2) ANN model resulted a coefficient of determination (R-2) of 0.95 for 1-decene conversion and 0.99 for 2-phenyldecane selectivity, while the R-2 of RSM was 0.93 and 0.92 for these two parameters. (C) 2013 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.