The Research On Analytical Method For Molecular Composition Of Diesel Distillate

Petroleum is an important resource for economic and social development.As a non-renewable energy source,petroleum is related to the lifeline of the national economy.In recent years,our country’s crude oil consumption has increased year by year,and its foreign dependence has increased year by year.In 2020,our country’s foreign dependence on crude oil has exceeded 73%.Efficient utilization of petroleum resources is an important issue currently faced by our country’s oil refining enterprises.Guided by molecular management technology and concepts,the technical level of petroleum processing can be raised and the utilization efficiency of petroleum can be improved.Thus,it is of great significance for reducing crude oil consumption,reducing the dependence on crude oil,and improving China’s oil safety.The petroleum molecular composition model is the basis for the application of molecular management technology,by providing accurate molecular information of oil products for process models such as molecular-level reaction kinetic models and molecular-level oil blending models.Based on this,the purpose of this paper is to establish a molecular composition analysis method for diesel fractions in an actual refinery,and to invert the molecular composition model based on the macroscopic physical properties of diesel fractions.The main research contents and conclusions are as follows:Reconstruction of diesel fraction molecules based on the structure-oriented lumping method was done,through expressing diesel molecules with 14 characteristic structures,and establishing a molecular library of diesel fractions based on more than 60 core molecules.The molecular library contains 678 typical molecules.The carbon number distribution is C8～C26,and the molecular physical property calculation is completed based on the group contribution method,and a diesel molecular information database containing molecular physical property information is constructed;A prediction model for the molecular composition of diesel fractions based on the maximum information entropy method was established.The PNA family composition,distillation range and density of diesel were used as the input of the model,and the molecular contents of 678 kinds of diesel fractions were used as the output.The comparison of 18 groups of sample data shows that the maximum information entropy method can greatly reduce the unknown variable parameters,simplify the calculation complexity,and predict the macroscopic physical properties of diesel fractions with high accuracy.The optimization effect proves that the average error of the predicted value of the particle swarm optimization algorithm is smaller than the average error predicted by the conjugate gradient method,and the simulated properties of the particle swarm optimization algorithm are closer to the properties of the actual samples,the optimization efficiency is higher,and it is more suitable for practical industrial applications.