Abstract:

Accurate prediction of reconstructed molecular fractions provides foundational data for constructing reaction kinetics models and simulating actual refining processes. And this contributes to enhance raw material utilization and product quality. Available molecular reconstruction models have paid more attention to the simulation accuracy of bulk properties rather than molecular composition. The reconstructed molecular library was moderately complex and well-organized by superposing core structures and homologous sequences. The molecular structural distribution patterns were studied. Constrains on molecular compositions were added and the molecular reconstruction model of oil composition based on molecular structural distribution was proposed. Vacuum gas oil molecular library was reconstructed with the reference of models that only focused on bulk properties. The reconstructed molecular structures and fractions were closer to actual oil, and the relative error of bulk properties was 1.08%.The results showed that the constrains of molecular structural distribution patterns could improve the simulation accuracy of molecular composition with little bulk properties error.

Key words: molecular simulation, model, optimization, molecular reconstruction, homologous series, vacuum gas oil

摘要：

准确的分子重构组成预测，为反应动力学模型的构建和实际炼油过程模拟提供基础数据，有助于提高原料利用率和产品品质。目前油品分子重构技术主要关注宏观物性的模拟精度，较少关注分子组成的预测准确性。本文利用核心结构叠加同源序列的方法，构造了层次清晰、复杂程度适中的分子库。探索油品中的分子结构分布规律，增加分子组成方面的约束条件，提出基于分子结构分布的油品组成分子重构模型。对减压蜡油进行分子重构，与仅关注宏观物性的模型相比，本研究重构得到的分子结构和组分含量更接近实际油品组成，同时宏观物性的平均相对误差为1.08%。研究结果表明，以分子结构分布规律为约束条件，可有效提高分子组成的模拟精度，同时保证宏观物性的误差较小。

关键词: 分子模拟, 模型, 优化, 分子重构, 同源序列, 减压蜡油