Using A Real-coded Genetic Algorithm to Predict Characterization Parameters for Petroleum Fraction

Abstract

In this paper, an efficient real-coded genetic algorithm (RCGA) for process optimization is developed to build new correlations to estimate properties of petroleum. This equation is a function of two input parameters that can be easily obtained. Moreover, an available data bank was used to develop the correlations using optimization. general experiment results reveal that the proposed RCGA is simple to use and provides a significantly faster convergence speed and much better search performance than comparative methods. The results of the proposed models are compared to others recommended in literature that have had large acceptance in the oil industry. The comparison results indicate that the proposed model is more precise than the most common models for characterizing petroleum fractions.