A neurocomputing model to calculate the thermo-solutal diffusion in liquid hydrocarbon mixtures

A thermodiffusion model based on the principles of artificial neural networks has been proposed. In addition to the development and validation with respect to the experimental data of binary n-alkane mixtures from the literature, the ability of the model to predict the known thermodiffusion trends has been presented. Comparison with other models in the literature is made to establish the superiority of the method. The model can be easily extended to multi-component mixtures and has potential to be applicable to study thermodiffusion in a wide variety of mixtures, viz., polymers, water-alcohol, colloidal mixtures, etc.