基于卷烟纸特性参数的中支烟主流烟气成分释放量预测

Prediction of the Release of Conventional Components in Medium-Sized Cigarette Smoke Based on the Characteristic Parameters of Cigarette Paper

为准确预测中支卷烟主流烟气中焦油、烟碱和CO释放量,使用随机森林算法构建了基于卷烟纸特性参数的焦油、烟碱和CO释放量预测数学模型,并进行了焦油、烟碱和CO释放量影响因素重要性排序和相关性分析。采用构建的随机森林算法模型对5个中支卷烟样品进行了焦油、烟碱和CO释放量的预测并与实测值进行了对比,验证模型的预测能力,同时对焦油、烟碱和CO释放量检测过程中可能产生的相对扩展不确定度进行测定。结果表明:基于卷烟纸特性参数构建的随机森林算法模型对于中支卷烟样品的焦油、烟碱和CO释放量的预测,具有较好的准确性、精确度和稳定性,且模型的焦油、烟碱和CO释放量预测残差均小于检测的扩展不确定度。

To accurately predict the release of tar,nicotine,and CO in the mainstream smoke of medium-sized cigarettes,a mathematical model for predicting tar,nicotine,and CO release based on cigarette paper characteristic parameters was constructed using the random forest algorithm.The importance ranking and correlation analysis of factors affecting tar,nicotine,and CO release was also conducted.The constructed random forest algorithm models were used to predict the release of tar,nicotine,and CO from 5 medium-sized cigarette samples and compared with the measured values to verify the predictive ability of the model.At the same time,the relative expanded uncertainty that may arise during the detection of tar,nicotine,and CO emissions was measured.The results showed that the random forest algorithm models constructed based on the characteristic parameters of cigarette paper has good accuracy and stability in predicting the release of tar,nicotine,and CO in medium cigarette samples,and the residual prediction of tar,nicotine,and CO release in the model is less than the expanded uncertainty of the detection.