Experimental study of natural gas hydrates and a novel use of neural network to predict hydrate formation conditions

► Importance of hydrate prediction and different predicting methods were explained. ► About 130 hydrate equilibrium points were obtained experimentally for 6 gas mixtures. ► Neural network accuracy in hydrates prediction was assessed by experimental data. ► Various architectures of neural networks and training functions have been attempted. ► The optimum structure and best variable definition was found and tested successfully.