电子鼻在小麦品质控制中的应用研究  

Application of electronic nose in the wheat quality control

作　　者：赵丹[1,2] 张玉荣[1] 林家永[2] 周显青[1] 

机构地区：[1]河南工业大学粮油食品学院,河南郑州450052 [2]国家粮食局科学研究院,北京100037

出　　处：《粮食与饲料工业》2012年第3期10-15,共6页Cereal & Feed Industry

基　　金：科技部"十一五"科技支撑项目(2009BADA0B00-5)

摘　　要：采用PEN3型电子鼻系统对我国10个省份47个小麦样品的挥发性物质进行检测。通过Loadings分析不同传感器在模式识别中的贡献率,对传感器阵列进行优化,并对传感器的响应值进行了主成分分析(PCA)和线性判别分析(LDA)。结果表明W5S、W1S、W2S3根传感器在小麦样品的用途、产地、品种区分识别时作用都较大,W1C、W3C2根传感器贡献率较小。PCA分析可以区分面包用小麦和馒头面条用小麦,总贡献率达85.5%;也可以区分不同产地的小麦样品和同一产地不同品种的小麦样品。线性判别分析(LDA)仅可以对不同产地、不同品种的小麦样品实现部分区分,无法将不同用途的小麦区分开来。主成分分析效果优于线性判别分析。The volatility of 48 wheat samples from 10 provinces in China were detected by electronic nose （PEN3）. The sen- sor array was optimized on the basis of the sensor contribution rates tested by the Loadings analysis. The sensors response data was analyzed by principal component analysis （PCA） and linear discrimination analysis （LDA）. The results showed that the sensors WSS, W1S,W2S had higher contribution rates during all the discrimination process, while W IC,W3C had lower rates. PCA analysis was able to identify wheat for bread and steamed bread or noodles making,with the contribution rates of 85.5 % and it was also able to identify samples from different producing areas. LDA could partly discriminate species or wheat samples from different producing areas,but it could not distinguish samples for different purposes. PCA discriminate effect was better than that of LDA.

关 键 词：小麦 挥发性物质 电子鼻 主成分分析 线性判别分析 Loadings分析 

分 类 号：S512.1[农业科学—作物学] TS211.2[轻工技术与工程—粮食、油脂及植物蛋白工程]