Study On Predicting Composition Of Diesel Oil Hydrocarbon In Molecular Level

The nature and structure of petroleum and its products has an important significance for oil refining process. Environmental regulation.is become more and more strict, for refining & chemical industry put forward higher request. This needs to make full use of petroleum resources. Therefore, thorough and accurate understanding the petroleum resources becomes a problem to be solved. Understanding of petroleum hydrocarbon molecules is one of key links. With the diesel as the research object, this paper carries out the research of predicting petroleum hydrocarbon molecules.First, this paper summarizes basic steps understanding petroleum hydrocarbons in the molecular level, introduces hydrocarbon component presentation template, the relationship of property and composition and optimization model in detail.Second, on the basis of oil molecular information database, it is put forward to a method of to building template of depending oncompound structure type and the carbon number distribution of diesel composition. According to the different main chain structure, analysis of the physical differences between diesel hydrocarbons structure type from the aspects of density, refractive index and the boiling point. Then it is established straight run diesel composition presentation template and Property template which contains 21 kinds of structure type and normal carbon number range. By using normal distribution and chi-square express diesel composition and the objective function.According to the characteristics of the diesel oil properties, established relationship of the physical properties between the monomer and macroscopic property. Using simulated annealing method to optimize objective function. According to group composition data and GC-TOF-MS data adjust the molecules composition. Eventually establish the diesel hydrocarbons molecular composition prediction model. At last, with samples of Straight-run diesel and catalytic diesel oil to verify the diesel hydrocarbons molecular composition prediction model Prediction results show that the model can well predict of straight run diesel hydrocarbons. Straight run diesel main components of the error between the predicted values and the measured values is less than 5%, Catalytic cracking diesel oil sample error is relatively a little big. But it can reflect the hydrocarbon distribution of the catalytic cracking of oil. Then the applicable scope of the forecast model is studied, it proves that the model has strong stability and extensive applicability to different diesel. It is able to achieve the requirement of the diesel fuel molecules prediction. Finally, the causes of error are analysed, it afforded the direction of the model improvement.