摘要
利用高光谱成像技术结合偏最小二乘(PLS)法建立不同成熟期贝达葡萄香气成分预测模型。用气相色谱-质谱(GC-MS)测定葡萄香气成分的组成及含量,选择葡萄香气的特征成分酮类、酚类、烷烃类为研究对象,通过连续投影算(SPA)法选择其特征波长,建立特征波长-香气成分模型和全波段-香气成分模型。结果表明:不同成熟期的葡萄香气成分含量与其光谱具有显著相关性:酮类物质全波段-PLS的预测相关系数(RP)为0.9925,SPAPLS的RP=0.9843;酚类物质全波段-PLS的RP=0.9888,SPA-PLS的RP=0.9928;烷烃类物质全波段-PLS的RP=0.9870,SPA-PLS的RP=0.9896。研究表明,高光谱成像技术可以对不同成熟期的葡萄香气成分进行无损检测。
Using the hyperspectral imaging technology and partial least square(PLS),a model was established to predict the aroma components of beta grapes with different maturity.The components and contents of grape aroma components were determined by gas chromatography-mass spectrometry(GC-MS).Ketones,phenols and alkanes,the characteristic components of grape aroma,were selected as the research objects.The characteristic wavelength was selected by continuous projection method(SPA),and the characteristic wave-aroma component model and whole band-aroma component model were established.The results showed that the content of aroma components in grapes at different maturity stages was significantly correlated with their spectra.The correlation coefficient(RP)of whole band-PLS was 0.9925,and SPA-PLS RP was 0.9843.RP of phenolic whole band PLS was 0.9888,and SPA-PLS was 0.9928.For alkanes,the RP of whole band-PLS was 0.9870,and SPA-PLS was 0.9896.The results showed that the hyperspectral imaging technology could be used to detect the aroma components of grapes with different maturity.
作者
靳雪梅
刘丽华
刘玉娇
郭月
毛迪锐
JIN Xuemei;LIU Lihua;LIU Yujiao;GUO Yue;MAO Dirui(Forestry College of Beihua University,Jilin 132013)
出处
《食品工业》
CAS
北大核心
2020年第11期324-328,共5页
The Food Industry
基金
2019年吉林省大学生创新项目(20190201096)。
关键词
高光谱成像技术
葡萄香气成分
无损检测
连续投影算(SPA)法
偏最小二乘(PLS)法
hyperspectral imaging technology
grape aroma component
nondestructive testing
successive projections algorithm(SPA)method
partial least squares(PLS)method
作者简介
通信作者:毛迪锐。
