Predicting Density, Viscosity,Volume and Temperature for Kerosene and Gas Oil Suplied by Doura Refinery From a Measured Experimental Result as Input Variables to Artificial Neural Network (ANN)

Abstract

Abstract: This study shows the possibility of predicting specific characteristicsof kerosene and gas oil using experimental measurements as inputs. Variables for programming (ANN) of artificial neuron networks. This study examines fuels from Dora refinery and investigates the relationship between fuel density and various properties such as volume, viscosity, and temperature. The correlation equation was determined by several linear regression analyses, and the coefficient (R2) showed some measurements in a strong achieved:R = 0.99986 for keosene, R= 0.99886 for gas oil.