Estimativa de parâmetros para a Equação Patel Teja por meio da otimização multiobjetivo com três objetivos

Abstract

The estimation of empirical parameters for the equations of state follows a classical approach, minimizing a function that aggregates the relative deviations of the saturation pressure and saturated liquid volume of the fluids of interest at different temperatures. This methodology, in addition to hiding the conflict in the adjustment of the two thermodynamic properties, neglects the deviation of the volume of the saturated gas and related functions, such as the enthalpy of vaporization. In this work, the empirical parameters of the Patel-Teja Equation were estimated with three objective functions that independently evaluate the deviation of the saturation pressure, the volume of the saturated liquid and the enthalpy of vaporization of pure substances. The resulting Pareto fronts were generated using a refinement technique with the imposition of restrictions on the codomain in conjunction with the multi-objective optimization method MOPSO-Light. The methodology applied allowed the visualization, the definition and increase in the density of points within the so-called “knee” regions. The optimal parameters chosen within these regions presented a better overall performance in predicting thermodynamic properties.