In complex scientific experiments, regression analysis has important value in data processing and analysis. Taking the preparation of C4 olefin by ethanol coupling as an example, this project established a regression model based on correlation analysis to solve the problem of the relationship between ethanol conversion, C4 olefin selectivity and temperature and time, and obtained that the goodness of fit of the quadratic fitting of ethanol conversion, C4 olefin selectivity and temperature was higher than that of the primary fitting, with an average of 0.03 higher. The selectivity of C4 olefin fluctuates between 36% and 41%, indicating that the selectivity of C4 olefin has little correlation with time under a given catalyst combination at 350℃. The ethanol conversion rate decreases with the increase of time, and finally tends to a constant value. According to the requirements of "Ethanol coupling preparation of C4 olefin" in Question B of the 2021 National Mathematical Competition in Modeling for College Students and the data provided in Annex 1 and Annex 2 of the question, this project establishes a multiple linear regression model and a single objective optimization model based on nonlinear regression to conduct a comprehensive study on ethanol conversion and maximum yield of C4 olefin.

Shucheng Zhong, Xuanning Wang, Qiang Chen. Application of regression analysis in the experiment of ethanol coupling preparation of C4 olefin. Academic Journal of Materials & Chemistry (2024) Vol. 5, Issue 3: 65-72. https://doi.org/10.25236/AJMC.2024.050310.

[1] Hongjie J, Yue Z, Liguo Z, et al. Stochastic pavement reconstruction method based on harmonic superposition technique of cubic spline interpolation[J]. Journal of Vibration and Shock, 2023, 42(10):23-30+143. 

[2] Haiqing C, Xu Z, Leilei Z, et al. Fitting generalized logistic distribution by least squares based on the logistic transformation of order statistics[J]. Communications in Statistics - Theory and Methods, 2023, 52(2). 

[3] Yuanshang Z, Weifang L. Research on typical scenarios based on Pearson correlation coefficient fusion density peak and entropy weight method[J]. Electric Power, 2023, 56(05):193-202. 

[4] Feng D, Li S, Hui, D, et al. Environmental correction of log curves based on multiple linear fitting and convolutional neural networks: A case study of Paleogene-Cretaceous strata in Xinhe-Sandaoqiao area[J]. Journal of Engineering Geophysics, 2023, 20(02):253-265. 

[5] Qi Y. An empirical study of influencing factors of China's foreign exchange reserves based on multiple linear regression model[J]. Theory of Chinese commerce, 2023(13):8-11.  

[6] Wei W, Meng W, Jun C, et al. Establishing a dual multivariate nonlinear regression constitutive equation for EB furnace melting TC4 titanium alloy based on DMNR model [J]. Rare Metal Materials and Engineering, 2021, 50(10):3609-3620.  

[7] Qiang Z, Weipao M, Qingsong L, et al. Optimization design of vertical axis wind turbine special airfoil based on multi-objective genetic algorithm[J]. Journal of Solar energy, 2023, 44(04):9-16.