The Prediction of Semiclathrate Hydrate Dissociation Conditions by Artificial Neural Network Tools for the Systems of (Methane / Carbon Dioxide / Nitrogen) + TBAC + Water

Semiclathrate hydrates are new structure of gas hydrates that dramatically promote the gas hydrate dissociation conditions, but there are few thermodynamic models to predict the semiclathrate hydrate dissociation conditions. In this research, the multi-layer perceptron artificial neural network tools were employed to predict the semiclathrate hydrate dissociation conditions for the systems of methane TBAC water, carbon dioxide TBAC water, and nitrogen TBAC water. A wide range of experimental data which was reported in the literature was used to develop this algorithm. Mass fraction (0 - 0.3618) TBAC aqueous solution data were utilized to train the artificial neural network. %85 of literature data points were used to train and develop the network and 15% of literature data points were used to examine the developed artificial neural networks. The predicted data by the developed artificial neural network (for the systems of methane TBAC water, carbon dioxide TBAC water, and nitrogen TBAC water) showed an acceptable agreement with experimental data.