Artificial Neural Network Modeling of a Reactive Distillation Process for Biodiesel Production

Abstract

The development of an artificial neural network model used to predict the mole fraction of biodiesel obtained from the reactive distillation process involving a reaction between palmitic acid and methanol has been carried out in this work. The data used for the neural network model development were generated with the aid of the parametric utility of Aspen HYSYS prototype of the process developed using Distillation Column Sub-Flowsheet and Wilson model as the fluid package. Six parameters (palmitic acid feed temperature, palmitic acid feed pressure, methanol feed temperature, methanol feed pressure, reboiler heat duty and reflux ratio) were used as the input variables to predict the biodiesel mole fraction. In the training of the neural network model, multiplicative lagged Fibonacci generator with the obtained optimized seed number of 47 was used. The results obtained from the training of the artificial neural network model developed revealed that the model was able to represent the process very well because the mean of squared error estimated from the simulation of the model was found to be approximately zero (that is, 3.93E-09). Furthermore, very good performance was recorded when the developed neural network model was used to predict the mole fraction of biodiesel produced from the bottom section of the reactive distillation column considered owing to the fact the mean of squared error was found, in this case, to be 6.92E-10. It has, thus, been discovered that the developed neural network model was able to perform very well in simulating and predicting the mole fraction of biodiesel produced from the reactive distillation process. Therefore, artificial neural network model has been shown to be a very good tool in representing a complex process like biodiesel reactive distillation.