为准确分析四川省烟草化学成分在时间和空间上的变异,构建新的综合赋权评价体系进行烟草化学成分可用性评价。以2010—2015、2017—2018年四川省5个市（州）的‘云烟87’上部烟叶316个样本的化学成分指标为数据集,利用隶属度函数对化学指标赋值,再通过主观、主成分、变异系数和综合赋权法,对每个指标重新赋权,从而评价烟草化学成分可用性。结果表明:（1）除钾和钾氯比2个指标外,其他绝大多数指标互为强相关。（2）四川省‘云烟87’上部烟叶还原糖、总烟碱含量偏高,氮含量和氮碱比偏低;还原糖含量处于最优范围的比例不高,但总氮、氮碱比、钾和钾氯比表现良好;整体上四川省烟草化学成分指标处于较适宜范围。（3）2010年和2017年化学成分可用性指数显著低于其他年份(P<0.05)。（4）凉山和攀枝花烟叶化学成分可用性指数明显高于宜宾(P<0.05),其中凉山表现更稳定。（5）综合赋权法相比于单一赋权法,虽然降低了烟叶化学成分可用性指数但却增大了样品间的差异性。相关性分析表明,组合赋权法结果更可靠。四川省‘云烟87’上部烟叶各化学成分指标表现良好,时空变异显著。综合赋权法增大了样品间化学成分可用性指数的变异,故更能可靠评价四川省烟草化学成分的组成特性。

To analyze the temporal and spatial variations of tobacco chemical composition usability in Sichuan Province, a new comprehensive evaluation system was established to calculate the usability index of tobacco chemical components. Taking the chemical indexes of 316 samples of ‘Yunyan 87’ upper leaves from five regions of Sichuan from 2010 to 2015 and from 2017 to 2018 as data sets, the chemical indexes were reassigned by membership function. Then, each index is reweighed by subjective, principal component, coefficient of variation and comprehensive weighting method, to evaluate the usability of tobacco chemical components. The results showed that: (1) except potassium and the ratio of potassium to chlorine, most other indexes were strongly correlated with each other; (2) the contents of reducing sugar and total nicotine of tobacco leaves in Sichuan were higher than standard contents, while the content of nitrogen and the ratio of nitrogen to alkali were lower; the proportion of reducing sugar content in the optimal range was small, but the total nitrogen, nitrogen alkali ratio, potassium and potassium chloride ratio performed well; on the whole, the coordination of chemical components of tobacco in Sichuan was good; (3) the usability index of tobacco chemical components in 2010 and 2017 was significantly lower than that in other years (P<0.05); (4) the usability index of tobacco chemical components in Liangshan and Panzhihua were significantly higher than that in Yibin, and that of Liangshan was more stable; (5) compared with the single weighting method, the comprehensive weighting method reduced the comprehensive score, but increased the difference between samples. The correlation analysis shows that the combination weighting-based comprehensive evaluation method is more reliable. The chemical components of upper leaves of ‘Yunyan 87’ in Sichuan Province perform well, and the spatial and temporal variations are significant. The comprehensive weighting method could increase the variation of chemical composition usability index among samples, so it is more reliable to evaluate the chemical composition characteristics of tobacco in Sichuan Province.