机器学习驱动的生物乙醇催化转化反应条件筛选  

Machine Learning-driven Intelligent Design of Experimental Conditions for Catalytic Conversion of Bioethanol

作　　者：夏薇[1] 张艳丽[1] 刘欣悦 蒋振华[1] 王馨蕊 陈坤[1] XIA Wei;ZHANG Yanli;LIU Xinyue;JIANG Zhenhua;WANG Xinrui;CHEN Kun(College of Chemistry and Chemical Engineering,China University of Petroleum(East China),Qingdao 266580,Shandong,China)

机构地区：[1]中国石油大学(华东)化学化工学院,山东青岛266580

出　　处：《实验室研究与探索》2025年第2期1-4,共4页Research and Exploration In Laboratory

基　　金：国家自然科学基金项目(21978327);山东省本科教学改革研究项目(M2023288);中国石油大学校级教学改革项目(CM2024040);中国石油大学校级研究生教育教学改革项目(YJG2022028);中国石油大学化学工程学院教学改革项目(YJ-K202003)。

摘　　要：为了将人工智能技术融入实验教学,设计了机器学习驱动的乙醇转化制备低碳烯烃催化剂实验,通过机器学习方法优化氧化锆催化生物乙醇制备丙烯的反应条件。引入随机森林模型和神经网络模型对催化剂煅烧温度、反应温度和反应压力3个关键反应条件进行分析。实验结果表明,催化剂煅烧温度550℃、反应温度470℃、反应压力10 atm(1 atm=101.325 kPa)条件下,氧化锆催化生物乙醇制备丙烯的效果最佳。此外,采用斯皮尔曼秩相关系数和肯德尔秩相关系数量化了3个反应条件对丙烯产率的影响权重。催化剂煅烧温度权重最大,反应温度次之、反应压力最小。In order to integrate artificial intelligence technology into experimental teaching,the experiment of catalyst for ethanol conversion to light olefin driven by machine learning was designed.The reaction conditions of catalytic conversion of bioethanol to propylene on zirconia were optimized by machine learning method.Three key reaction conditions of calcination temperature,reaction temperature and reaction pressure were analyzed by introducing random forest model and neural network model.The experimental results show that zirconia has the best effect on preparing propylene from bioethanol when the calcination temperature is 550℃,the reaction temperature is about 470℃,and the reaction pressure is about 10 atm(1 atm=101.325 kPa).In addition,Spearman rank coefficient and Kendall rank correlation coefficient were used to quantify the influence weights of three reaction conditions on propylene yield.It was found that the weight of calcination temperature is the largest,followed by reaction temperature and reaction pressure.

关 键 词：实验设计 机器学习 氧化锆 生物乙醇 丙烯制备 

分 类 号：G642.0[文化科学—高等教育学] TQ426.94[文化科学—教育学] G642.423[化学工程]