A Novel Model for Predicting Lower Flammability Limits Using Quantitative Structure Activity Relationship Approach

摘要：

　　The lower flammability limit (LFL) of a chemical is the primary property for assessing the fire hazards of a chemical in its vapor or gas state.Therefore it is the crucial information for people handling flammable gas for safety concerns.In this work, a quantitative structure activity relationship (QSAR) model for predicting the LFL of organic compounds is developed based on 458 experimental measurements of LFL data.The stoichiometric concentration, and other molecular descriptors provided by CODESSA PRO software and Dragon 6 were adopted as the candidate regressors of a multiple regression model.Using a modified stepwise regression algorithm, a four-descriptor model is suggested.This model gives a goodness-of-fit performances with the R2 of 0.956, and a predictive performance with the Q2 of 0.914.The application domain for the proposed model also been carefully discussed.The average absolute error of training set and test set for the proposed model are found to be of 0.341 vol % and 0.269 vol %, respectively.Compared with the experimental uncertainty for determining LFL experimentally, the proposed QSAR model gives reasonable performance.

关键词： Lower flammability limit Quantitative Structure Activity Relationship Model

作者: Yueh-Chun GUO Chan-Cheng CHEN Chao-Pei LAI

作者单位: Department of Safety, Health and Environmental Engineering, National Kaohsiung First University of Science and Technology,No.1 University Road, Yanchao District, Kaohsiung 824, Taiwan, China

会议类型: 国际会议

会议名称: The 2016 International Symposium on Safety Science and Technology(2016安全科学与技术国际会议)

会议地点: 昆明

会议语种:英文

页码: 419-425

在线出版日期: 2016-10-17（万方平台首次上网日期，不代表论文的发表时间）