叶组配方卷烟烟气预测模型的建立

Models for Prediction of Mainstream Cigarette Smoke of Tobacco Blend

利用多元回归分析方法建立了叶组配方焦油量、烟气烟碱量、一氧化碳量和抽吸口数的预测模型,并筛选出了与这4项指标关系密切的化学成分,包括总糖、还原糖、总氮、烟碱、挥发碱、钾、氯、硫酸根等。利用所建模型对4种卷烟产品叶组配方的烟气指标进行了预测,结果表明,其焦油量、烟气烟碱量、一氧化碳量和抽吸口数的预测误差范围分别为0.04～1.13mg/支、0.13～0.50mg/支、0.11～2.95mg/支和0.37～1.15口/支。

The models for predicting the tar, nicotine, CO yields in mainstream cigarette smoke were developed with multiple regression analysis, the key constituents in tobacco closely related to the mentioned three compounds were screened out, including total sugar, reducing sugar, total nitrogen, nicotine, volatile base, potassium, chlorine and sulfate. The tar, nicotine and CO yields in cigarette smoke of four cigarette blends were estimated with these models. The results showed that the absolute estimate errors of tar, nicotine and CO yields were in the ranges of 0.04 - 1.13, 0. 13 - 0.50, and 0. 11 - 2.95mg/cig, respectively.