基于机器学习的硅胶泡沫复合材料LOI预测研究  

LOI Prediction of Silicone Rubber Foam Composites Based on Machine Learning

作　　者：刘博 宫花 冀贤 LIU Bo;GONG Hua;JI Xian(College of Safety Science and Engineering,Xi'an University of Science and Technology,Xi'an Shaanxi 710054,China)

机构地区：[1]西安科技大学安全科学与工程学院,陕西西安710054

出　　处：《当代化工》2024年第11期2521-2525,2549,共6页Contemporary Chemical Industry

基　　金：国家自然科学基金项目(项目编号:51904233)。

摘　　要：为了探究机器学习(ML)预测硅橡胶泡沫(SiFs)阻燃材料极限氧指数(LOI)的可行性,通过建立改性蒙脱土(MMT)/硅橡胶阻燃泡沫材料LOI数据库,分析了影响LOI的工艺参数的重要性,采用线性回归(LR)、支持向量回归(SVR)、决策树(DT)、随机森林(RF)、极致梯度提升(XGBoost)5种机器学习算法搭建预测模型,并通过网格搜索和交叉验证对精度较高的模型超参数进一步优化训练。结果表明:工艺参数对SiFs的LOI预测模型贡献程度由大到小顺序为阻燃剂、抑制剂、20~30 mPa·s羟基硅油、反应温度、催化剂、含氢硅油。阻燃剂含量与LOI预测模型的皮尔逊相关系数最大绝对值为0.58。XGBoost模型最适合用于预测LOI,判定系数(R2)最高到0.811。研究结果为缩短SiFs阻燃材料的研发周期提供新思路。In order to explore the feasibility of machine learning(ML)in predicting the limiting oxygen index(LOI)of silicone rubber foam(SiFs)flame retardant materials,the importance of process parameters affecting LOI was analyzed by establishing the LOI database of modified montmorilonite(MMT)/silicone rubber flame retardant foam materials.Five machine learning algorithms,linear regression(LR),support vector regression(SVR),decision tree(DT),random forest(RF)and extreme gradient boosting(XGBoost),were used to build the prediction model,and the hyperparameters of the model with high accuracy were further optimized and trained by grid search and cross-validation.The results showed that the degree of contribution of process parameters to the LOI prediction model of SiFs was in the order of flame retardant>inhibitor>20~30 mPa·s hydroxyl silicone oil>reaction temperature>catalyst>hydrogen-containing silicone oil.The absolute value of Pearson correlation coefficient between flame retardant content and LOI prediction model was 0.58.The XGBoost model was most suitable for predicting LOI values,with the determination coefficient(R2)up to 0.811.The research results provide a new way to shorten the R&D cycle of SiFs flame retardant materials.

关 键 词：机器学习 硅橡胶泡沫 极限氧指数 阻燃性能 极致梯度提升 

分 类 号：TQ333.93[化学工程—橡胶工业]