基于QSPR的烃类物质最大实验安全间隙理论预测研究

Theoretic Prediction Study of Maximum Experimental Safe Gap of Hydrocarbons Based on Quantitative Structure Property Relationship

从分子结构角度出发,应用定量构效(QSPR)方法对烃类物质最大实验安全间隙进行系统研究。采用国际电工委员会标准数据作为样本集,应用遗传算法实现分子结构特征描述符选择和优化,基于MLR线性方法建立化合物性质与其分子结构参数间的定量函数关系模型,得到烃类物质最大实验安全间隙理论预测模型,并对模型进行验证和应用域分析。结果表明,最大实验安全间隙的预测值与实验值符合良好,预测模型具有较好稳定性和泛化推广性能。相关研究为工程上提供了一种根据分子结构预测最大实验安全间隙的新方法。

From the perspective of molecular structure,the quantitative structure property relationship(QSPR)method was used to study the maximum experimental safety gap(MESG)of hydrocarbons.The data of International Electrotechnical Commission was chosen as sample set,characteristic descriptors of molecular structure were selected and optimized based on Genetic Algorithm(GA).Multiple linear regression method was employed to establish prediction model of MESG.Subsequently,the validation and interpretation were performed and the Williams plot was drawn to analyze the applicability domain.The results show that the predicted value of the maximum experimental safety clearance is in good agreement with the experimental value,and the prediction model has good stability and generalization performance.The related research provides a new method for engineering to predict the maximum experimental safety gap based on molecular structure.