Prediction of hydrocarbon densities using an artificial neural network–group contribution method up to high temperatures and pressures

In this work, the densities of hydrocarbon systems have been estimated using a combined method that includes an artificial neural network (ANN) and a simple group contribution method (GCM). A total of 2891 data points of density at several temperatures and pressures, corresponding to 40 different hydrocarbons including short- and long-chain alkanes ranging from CH4 to n-C40H82, and also several cycloalkanes, highly branched alkanes and aromatic hydrocarbons have been used to train, validate and test the model. This study shows that the ANN–GCM model represent an excellent alternative for the estimation of the density of hydrocarbons with a good accuracy. A wide comparison between our results and those of obtained from some previous methods shows that this work can provide a simple procedure for prediction the density of different classes of hydrocarbons in a better accord with experimental data up to high temperature, high pressure (HTHP) conditions.

► The densities of hydrocarbons have been estimated using an ANN–GCM method up to HTHP.
► The best network configuration consisted of 21 neurons in the hidden layer.
► The advantage of this technique is its high speed, simplicity and generalization.
► A wide comparison between this method and some previous works has been made.
► The AAD for train, validation, and test sets are 0.30, 0.34, and 0.39, respectively.