Feed Forward Neural Network Model for Isopropyl Myristate Production in Industrial-scale Semi-batch Reactive Distillation Columns

• Main Authors: Bashah, Nur Alwani Ali, Othman, Mohd Roslee, Aziz, Norashid
• Format: Article
• Language: English
• Published: Taylor's University 2015
• Subjects: TA1-2040 Engineering (General). Civil engineering (General)
• Online Access: http://eprints.usm.my/42785/1/JES_Vol._11_2015_-_Art._6%2859-65%29.pdf; http://eprints.usm.my/42785/; http://web.usm.my/jes/11_2015/JES%20Vol.%2011%202015%20-%20Art.%206(59-65).pdf
• Institution: Universiti Sains Malaysia

The application of the artificial neural network (ANN) model in chemical industries has grown due to its ability to solve complex model and online application problems. Typically, the ANN model is good at predicting data within the training range but is limited when predicting extrapolated data. Thus, in this paper, selected optimum multiple-input multiple-output (MIMO) and multiple-input single-output (MISO) models are used to predict the bottom (xb) compositions of extrapolated data. The MIMO and MISO models both managed to predict the extrapolated data with MSE values of 0.0078 and 0.0063 and with R2 values of 0.9986 and 0.9975, respectively.