回归分析在乙醇偶合制备C4烯烃实验数据分析及模型参数拟合中的应用

Application of Regression Analysis for Data Analysis and Model Parameter Fitting in Experiments of C4Olefin Preparation by Ethanol Coupling

回归分析方法在复杂科学实验的数据分析处理方面具有重要价值,本文以乙醇偶合制备C4烯烃为案例,基于回归分析与线性规划模型,从数理统计角度深入分析了乙醇催化偶合制备C4烯烃的工艺条件。通过分析温度与乙醇转化率和C4烯烃选择性的相关性,针对不同催化剂组合建立了基于皮尔逊相关性分析的多项式回归方程的模型;在此基础上,分析多变量对C4烯烃收率的影响,并通过对多元线性回归方程的求解,建立单目标线性规划模型并获得其最优解。

Regression analysis method is of great importance for data analysis and processing in the fields of complex scientific experiments.In this paper,we report the use of regression analysis and linear programming model for systematically analyzing the experimental data and assessing the optimal operating conditions for the preparation of C4 olefin by ethanol coupling reactions.Through analysis of the correlations between reaction temperature,ethanol conversion rates and C4 olefin selectivity,a polynomial regression equation model based on Pearson correlation was established at first with respect to the various combinations of different catalysts.On this basis, the influence of multiple variables on C4 olefin yield was analyzed, and the single-objective linear programming model was established and its optimal solution was obtained by solving the multiple linear regression equation.Using the interpolation and fitting algorithm,we established a single objective linear programming model and obtained its optimal solution through solving the multiple linear regression equation.