山东烤烟感官质量与化学成分关系及质量预测模型建立  

Relationship Between Sensory Quality and Chemical Component of Flue-cured Tobacco in Shandong Province and Establishment of Quality Prediction Models

作　　者：周显升[1] 李晓阳 刘志广[1] 周小雨[1] 邱承宇[1] 庄志麟 曹建敏[3] ZHOU Xiansheng;LI Xiaoyang;LIU Zhiguang;ZHOU Xiaoyu;QIU Chengyu;ZHUANG Zhilin;CAO Jianmin(China Tobacco Shandong Industry Co.,Ltd.,Jinan 250013;Qingdao University,Qingdao,Shandong 266071;Tobacco Research Institute,Chinese Academy of Agricultural Sciences,Qingdao,Shandong 266101;Graduate School,Chinese Academy of Agricultural Sciences,Beijing 100081)

机构地区：[1]山东中烟工业有限责任公司,济南250013 [2]青岛大学,山东青岛266071 [3]中国农业科学院烟草研究所,山东青岛266101 [4]中国农业科学院研究生院,北京100081

出　　处：《中国农学通报》2024年第1期128-134,共7页Chinese Agricultural Science Bulletin

基　　金：中国农业科学院科技创新工程“中国农业科学院-烟草风味化学创新团队”(ASTIP-TRIC07);山东中烟工业有限责任公司重点项目“低适用片烟溶洗改性技术研究”(2021370000340022);中国农业科学院基本科研业务费所级统筹项目“烤烟中重要潜香型物质Amadori化合物含量特征和热解产香规律”(161023202100)。

摘　　要：为明确山东烤烟化学成分与感官质量之间的关系,探明影响山东烟叶质量提升的关键化学成分指标,对山东省6个地区代表性烟叶样品开展感官质量评价和常规成分、有机酸、生物碱、单双糖、多酚等5类23种化学指标测定,结合简单相关、方差分析、主成分分析和回归分析等统计方法进行数据分析。筛选出总糖、氯、绿原酸、降烟碱、柠檬酸/烟碱、油酸6个影响山东烟叶品质的关键化学指标,基于6个指标构建山东烟叶感官质量预测模型,验证样本预测值与真实值绝对差值0.43~3.88分,相对误差-5.66%~6.32%,平均误差3.18%。基于关键化学成分构建的感官质量预测模型可为山东烟叶质量的客观评价提供一定的技术支撑。In order to explore the relationship between chemical components and sensory quality of flue-cured tobacco in Shandong Province,a sensory quality evaluation and determination of 23 chemical indicators in five categories,including conventional components,organic acids,alkaloids,monosaccharides and polyphenols,were conducted on representative flue-cured tobacco leaves from six regions in Shandong Province.Multiple statistical methods such as simple correlation,analysis of variance,principal component analysis,and regression analysis were used for data analysis.Six key quality indicators including total sugar,chlorine,chlorogenic acid,nornicotine,citric acid/nicotine,and oleic acid were screened.The sensory quality prediction model of Shandong tobacco leaves was constructed.The absolute difference between the predicted value and the actual value was 0.43-3.88 points.The relative error was-5.66%-6.32%,between actual value and predicted value of the verification sample,and the average error was 3.18%.The sensory quality prediction model based on key chemical components can provide certain technical support for the objective quality evaluation of flue-cured tobacco in Shandong Province.

关 键 词：山东烤烟 多元统计分析 质量评价 化学成分 预测模型 

分 类 号：TS411[农业科学—烟草工业]