吉首大学学报(自然科学版) ›› 2022, Vol. 43 ›› Issue (3): 79-83.DOI: 10.13438/j.cnki.jdzk.2022.03.013

• 化学化工 • 上一篇    下一篇

基于随机森林与L-M算法的C4烯烃制备优化模型

蒋梓浩,任涛,周祺,骆加冕   

  1. (东北大学软件学院,辽宁 沈阳 110169)
  • 出版日期:2022-05-25 发布日期:2022-09-06
  • 通讯作者: 任涛(1980—),男,辽宁沈阳人,东北大学软件学院教授,博士,博士生导师,主要从事大数据分析及应用研究.
  • 基金资助:
    中央高校基本科研业务费资助项目(N181706001,N2017009);辽宁省自然科学基金国家重点实验室资助项目(2020-KF-12-11)

Optimization Model for C4 Olefin Preparation Based on Random Forest and L-M Algorithm

JIANG Zihao,REN Tao,ZHOU Qi,LUO Jiamian   

  1. (School of Software,Northeastern University,Shenyang 110169,Liaoning China)
  • Online:2022-05-25 Published:2022-09-06

摘要:探究合适的催化剂组合和温度对乙醇偶合制备烯烃有着重要意义.首先,利用方差分析筛选重要影响因子;然后,通过随机森林算法分析出影响因子对乙醇转化率、C4烯烃选择性的重要程度排序:温度>催化剂质量>乙醇进样量>Co负载量>装料比;最后,利用L-M算法进行多元非线性回归,以C4烯烃收率最优为目标并借助粒子群算法得到C4烯烃收率最优值为42.5485%.

关键词: 随机森林, L-M算法, C4烯烃制备, 粒子群算法

Abstract: It is important to investigate the catalyst combination and temperature on the preparation of olefins by ethanol coupling.Firstly,ANOVA was used to screen the important influencing factors;then,the importance of the influencing factors on ethanol conversion and C4 olefin selectivity was analyzed by random forest algorithm:temperature>catalyst mass>ethanol concentration>Co loading>charging ratio;finally,multiple non-linear regression was performed by L-M algorithm,and the optimal C4 olefin yield was obtained with particle swarm algorithm.The optimal value was 42.548 5%.

Key words: random forest, L-M algorithm, C4 , olefin preparation, particle swarm algorithm

公众号 电子书橱 超星期刊 手机浏览 在线QQ