Molecular Reconstruction Of Heavy Petroleum Fraction Based On Calibration Data Of Crude Oil

Molecular reconstruction is a method of building molecular mixtures from analytical data.In this article,based on the calibration data of crude oil,VGO and VR fraction of heavy petroleum fraction are reconstructed by stochastic reconstruction-entropy maximization method.First,properties estimation equations for petroleum molecules in heavy petroleum fraction are selected through verification calculation in a large scale.Next,the petroleum molecules are represented by the structure vectors and the characteristic value of the structure vectors are optimized by the genetic algorithm,the virtual molecular libraries which can reflect the real composition of VGO and VR fractions are obtained respectively by stochastic reconstruction method.Finally,the objective function is constructed based on the entropy maximization principle and constraint conditions are selected based on the calibration data of crude oil,and then the objective function is optimized by BGFS method to obtain mole fractions of the virtual molecular libraries of VGO and VR fraction and complete the molecular reconstruction of heavy petroleum fraction.The results show that molecular reconstruction can accurately simulate the molecular composition of heavy petroleum fractions from different crude oils.The research results can be applied to petroleum reaction kinetics and petroleum blending.