Prediction of Undersaturated Crude Oil Density Using Gaussian Process Regression

Abstract

Gaussian process is a powerful tool to model sophisticated tasks in the machine learning field. On the other side, density of crude oil is an important property in simulation processes and design of equipments. Nevertheless; using laboratory methods to measure crude oil density is costly and time consuming; thus, development of a predictive model to estimate the density of crude oil is very beneficial. The authors develop a Gaussian process–based model to predict the density of undersaturated crude oil. Results were compared with the previous works and it was shown that the new method outperforms them.