首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于主成分聚类分析的湘南烟叶子外观质量分类评价
引用本文:郭亮,;文芸,;杨红武.基于主成分聚类分析的湘南烟叶子外观质量分类评价[J].湖南农业科学,2014(7):53-56.
作者姓名:郭亮  ;文芸  ;杨红武
作者单位:[1]湖南省烟草公司长沙市公司,湖南长沙410003; [2]常德市烟草公司桃源县分公司,湖南常德415700; [3]湖南农业大学农学院,湖南长沙410128
基金项目:国家烟草专卖局浓香型特色优质烟叶开发项目(Ts-01);湖南浓香型特色优质烟叶开发项目(11-14Aa01);湖南浓香型特色优质烟叶研究与开发项目(2013NK3073)
摘    要:采用湘南烟区2007-2011年烟叶外观质量指标数据,运用专家咨询法、因子分析法、模糊数学等方法,建立了基于主成分分析和聚类分析的烟叶质量评价模型,并对湘南烟区14个县市84个烟叶样品进行烟叶外观质量分类评价。结果表明:2007-2011年湘南烟区烟叶外观质量指标变异系数小,年度间差异只有2008年偏大;成熟度是影响烟叶外观质量的最主要因子,占总权重的20%;84个样品的两个主成分聚类与主成分总分聚类结果相差不大,而湘南烟区烟叶外观质量集中在第2类烟(占32.14%),其次是第3类烟(占29.76%),而第5类烟最少(占3.57%)。

关 键 词:烟叶  外观质量  主成分分析  聚类分析  分类

Classified Assessments of Tobacco Leaf Appearance Quality Based on Principal Component Clustering Analysis in South Hunan
Institution:GUO Liang, WEN Yun, YANG Hong-wu(1. ChangSha company of Hunan Tobacoo company, Changsha 410128; 2.Taoyuan branch company of Chang-de tobacco company, Changde 415700; 3.Hunan Agriculture University, Changsha 410128, PRC)
Abstract:By using expertconsultation method,factoranalysisand fuzzy mathematics,a quality assessmentmodeloftobacco leavesbased on principalcomponentanalysis and clustering analysis was constructed according to appearance quality index data oftobacco area in South Hunan during 2007~2011,and then a classified assessment of appearance quality of tobacco leaves was conducted among 84 leaf samples from14countiesandcitiesinSouthHunan.ResultsshowedthatthecoefficientofvariationofSouth-Hunantobaccoareaduring2007~2011 wassmall,andtherewasarelativelyhugedifferencebetweenyearsonlyin2008.M aturityisamainfactorthataffectstheappearancequality oftobacco leaves,which occupies a proportion of20% .N o obvious differences were found between two principalcomponentclustering and totalclustering,and the appearance quality oftobacco leavesin South-Hunan tobacco area much belonged to the second category ofcigarette (about32.14% ),nextto the third category (about29.76% ),and leastto the fifth (about3.57% ).
Keywords:tobacco  appearance quality  main component analysis  clustering analysis  classification
本文献已被 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号