摘 要: 依据某石化企业催化裂化汽油精制脱硫装置运行四年产生的大量数据和从催化裂化汽油精制装置采集的325 组样本(每组样本中都有367 个操作变量),针对辛烷值损失的优化问题,从数学建模的方向出发,在对样本数据进行预处理的基础上,使用二次特征筛选的方法从367 个操作变量中筛选出数学建模的主要变量,综合考虑变量之间的非线性和相互强耦联性,基于多变量自回归对数线性方程建立了汽油精制过程中的辛烷值损失预测模型。 |
关键词: 辛烷值;损失预测模型;多变量自回归对数线性方程 |
中图分类号: TP183
文献标识码: A
|
|
Research on Octane Number Loss based on Vector Autoregressive Model |
QIN Qingtao1, GU Haihang2
|
( 1.School of Mechanical Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China; 2.School of Mechanical Engineering, University of Yancheng Institute of Technology, Yancheng 224000, China )
522636581@qq.com; 2237298110@qq.com
|
Abstract: This paper is based on a large amount of data generated by a certain petrochemical enterprise's catalytic cracked gasoline refining and desulfurization unit during its four-year operation, and 325 sets of samples collected from the catalytic cracked gasoline refining unit, with each group of samples having 367 operating variables. Aiming at the optimization of octane number loss, from the perspective of mathematical modeling, the secondary feature screening method is used to screen out the main variables of mathematical modeling from 367 operational variables on the basis of preprocessing the sample data. After comprehensive consideration of the non-linearity and strong mutual coupling between the variables, a prediction model of octane number loss in gasoline refining process is established based on multivariate autoregressive log-linear equation. |
Keywords: octane number; loss prediction model; multivariate autoregressive log-linear equation |