QSPR在混合物典型燃爆特性预测中的研究进展

Research progress of quantitative structure-property relationship(QSPR)method for predicting typical combustion and explosion characteristics of mixtures

为全面了解定量结构-性质关系(QSPR)方法在混合物燃爆特性预测中的研究现状,展望其发展趋势,综述其在混合物闪点、爆炸极限与自燃温度预测中的国内外研究进展,分析预测目标参数的选择、数据收集、描述符计算和筛选以及模型建立和验证等方面的不足与研究方向。结果表明:QSPR在混合物燃爆特性预测中尚处于起步阶段,当前研究的首要限制是混合物燃爆特性参数实验数据样本不足,关键点及难点是混合物结构的准确表征,未来研究应关注的重点是大量数据源统一的数据样本的获取方法、非加和性混合物分子描述符的计算方法以及机器学习等非线性建模方法。

To comprehensively understand the research status of quantitative structure-property relationship(QSPR)method for predicting the combustion and explosion characteristics of mixtures,and forecast its development trend,the research progress of QSPR method for predicting the flash points,flammability limits and auto-ignition temperatures of mixtures at home and abroad was reviewed and summarized.The deficiencies and research trend about the selection of prediction properties,the data collection,the calculation and selection of molecular descriptors,and the establishment and verification of model were analyzed.The results showed that QSPR method was still at an initial stage for predicting the combustion and explosion characteristics of mixtures.The primary limitation of current research was the lack of experimental data about the combustion and explosion properties of mixtures.The key point and difficult point was the accurate molecular descriptors of mixture structures.The focus of future research should be the acquisition method of a large number of experimental data with unified data source,the calculation methods for non-additive molecular descriptors of mixtures,and the Machine Learning and other nonlinear modeling methods.