Inquiry of C4 Olefin Chemical Technology Based on Statistical Modeling

Abstract

A large number of by-products will occur during the transformation of ethanol to C4 olefin, resulting in the selection of the destination products that are not enough to meet industrial needs. Therefore, the selection of catalyst combinations and environmental conditions are of great significance for the preparation of C4 olefin. First of all, this paper adopts a random forest model to screen the experimental characteristics and the Grid Search is used for optimizing hyperparameters to evaluate the importance of different catalyst types to ethanol conversion rate and C4 olefin selectivity. On this basis, with the multiple regression equations established among different catalyst types, temperature and C4 olefin income as the target function, under the constraint of temperature, Co loads, Co/SiO2 and HAP quality ratio, ethanol concentration, obtain the combination of catalysts with the highest income of C4 olefin. The proportion of C4 olefin in the target product was increased. At the same time, it provides reference for exploring suitable chemical process conditions of C4 olefin.