Estimation of thermophysical properties of dimethyl ether as a commercial refrigerant based on artificial neural networks

In this research, the ability of Multilayer Perceptron Artificial Neural Networks based on back-propagation algorithm has been investigated to estimate dimethyl ether (RE 170) densities and vapor pressures. The best network configuration for this case was determined as a three layers network including 12, 15, 1 neurons in its layers, respectively, using Levenberg–Marquardt training algorithm. The comparisons between results show that there is a good agreement between experimental data and network predictions. As the uncertainties in the presented network for prediction of and saturated liquid, densities are less than 0.04% and 0.16%, respectively. Comparisons among the ANN predictions, several equations of state, and experimental data sets show that the ANN results are in good agreement with the experimental data better than EoSs. Another network for estimation of vapor pressure has been trained with uncertainty less than 0.85%. Results prove that artificial neural network can be a successful tool to represent thermophysical properties effectively, if develops efficiently.