乙醇偶合制备C4烯烃的数学模型

Mathematical Model of C4 Olefin Preparation by Ethanol Coupling

针对2021年全国大学生数学建模竞赛B题“乙醇偶合制备C4烯烃”中的第3问和第4问,在分析已有解决方案的基础上,建立BP神经网络模型和Logit变换一次回归模型,通过赤池信息量准则等,比较并分析两种模型的结果,得到在一定范围内C4烯烃收率尽可能高的催化剂组合和温度.并运用均匀试验设计法,分别以C4烯烃收率最高时所对应的催化剂组合与温度取值的邻域及已给出的109组实验数据的变化区间作为试验区域,将L_(2)-星偏差作为均匀性度量标准,用穷举法求解,得到追加的5次最佳实验.

Aiming at the third and fourth questions in the 2021 CUMCM problem B"ethanol coupling to C4 olefins",based on the analysis of the existing solutions,this paper establishes the back propagation neural network model and the logit transformation one-time regression model,compares and analyzes the results of the two models through Akaike information criterion,etc.,and then obtains the catalyst combination and temperature with the highest C4 olefin yield within a certain range.And using the uniform experimental design method,taking the field of catalyst combination and temperature value corresponding to the highest C4 olefin yield and the variation range of 109 groups of experimental data as the test area,taking L_(2) star deviation as the standard of uniformity measurement,and then using the exhaustive method to obtain the required five best experiments.