Parametric Study of Methanol Reformer Membrane Reactor for Hydrogen Production Using Artificial Neural Network

Document Type: Original Article


1 Department of Nanomaterials, Faculty of Engineering, University of Tarbiat Modares, Tehran, Iran

2 Department of Chemical Engineering, Faculty of Engineering, University of Semnan, Semnan, Iran

3 HSE Department of Abadan Institute of Technology Petroleum University of Technolgy, North Bowardeh, Abadan, Iran


Efforts to find eco-friendly fuels have attracted researchers' attention to hydrogen and its production methods. In the present study, hydrogen recovery in a membrane reactor with a Pd-Ag catalyst for hydrogen production was simulated. For this purpose, the multilayer perceptron neural network (MLP) and radial basis function (RBF) were used. Moreover, implementing intelligent methods for simulation require a dataset, so 60 datasets were applied. The effects of parameters such as temperature, pressure and dimensionless length of reactor on the hydrogen recovery in the membrane reactor were investigated. The results show that the MLP method had a high adaptation to experimental data and had the best performance in prediction. Intelligent techniques can be used to reduce the computational time and increase the accuracy of the techniques. The values of the RMSE and R2 evaluation criteria for the best model are 0.3274 and 0.9997, respectively.