辛烷值损失预测模型的构建与优化

Construction and Optimization of Octane Loss Prediction Model

汽油是小型车辆的主要燃料,汽油燃烧产生的尾气排放对大气环境有重要影响。针对我国汽油在精制过程中辛烷值损失的问题,首先使用Lasso回归对工业样本数据进行预处理,再运用多元线性回归建立辛烷值损失预测模型,最后基于数据包络分析法(DEA)对汽油精制过程中的主要操作变量进行优化。该优化模型具有实际应用意义,为汽油清洁化过程降低辛烷值损失的研究提供了理论依据。

Gasoline is the major fuel for small-sized vehicles, and the gasoline-fueled vehicle emissions generated by gasoline combustion exert an important impact on the atmosphere. Regarding the loss of octane value amid the process of gasoline refining in our country, t Lasso regression is used to process the industrial specimen data beforehand, and then a prediction model for octane loss is established through multiple linear regression. Finally, the major operating variables used in gasoline refining are optimized based on Data Envelopment Analysis (DEA) model. The optimized model is of practical significance for application and proffers theoretical ground for the reduction of the loss of octane value during the refining of gasoline.