基于主成分与聚类分析的不同粉路小麦次粉品质综合评价

为了获取食品化利用品质高的小麦次粉,采用主成分分析与聚类分析方法对不同粉路小麦次粉进行分析和筛选。通过测定12个不同粉路小麦次粉和2个不同等级小麦粉的营养品质、色泽、面筋特性和糊化特性,应用主成分分析和聚类分析方法,筛选出不同粉路小麦次粉品质评价的主要指标并进行分类。结果表明:不同粉路小麦次粉的品质特性中,脂肪、灰分、面筋指数、峰值粘度、最低粘度、最终粘度、衰减值、回生值变异系数较大,均超过20%。利用主成分分析方法提取其中更具代表性的4个主成分因子,累计方差贡献率达到84.182%,依据综合评价函数进行分析计算排名,综合排名前三位的为DF1XF1下、DF1XF1上和精制面粉;聚类分析将不同粉路小麦次粉分为4类,次粉DF1XF1下和DF1XF1上与精制面粉归为一类,与主成分分析结果一致。12个不同粉路小麦次粉中,DF1XF1下和DFXF1上品质最好,7M下、8M上和2T下品质最差。基于主成分和聚类分析方法综合评价小麦次粉品质特性,可为其食品化高值利用提供参考。

Comprehensive Quality Evaluation of Different Wheat Middlings by Principal Component Analysis and Cluster Analysis

In order to obtain high-quality wheat middlings for food use, principal component analysis and clustering analysis were used to analyze and screen different varieties of wheat middlings. The nutritional quality, color, gluten characteristics and gelatinization properties of 12 different varieties of wheat middlings and 2 different grades of wheat flour were determined. The main indexes for quality evaluation of different varieties of wheat middlings were screened and classified by principal component analysis and cluster analysis. The results showed that among the quality characteristics of different wheat middling flours, the coefficients of variation (CVs) of the lipids, ash, gluten index, peak viscosity, trough viscosity, final viscosity, breakdown and setback values were relatively high (all>20%). By principal component analysis, 4 more representative principal component factors were extracted, the cumulative variance contribution rates reached 84.182%. Analysis, calculation and ranking were conducted according to the comprehensive evaluation function, developed based on principal component analysis, the top three of the comprehensive ranking were DF1XF1-under, DF1XF1-upper and refined wheat flour, respectively. The cluster analysis divided the different varieties of wheat middling samples into 4 categories, and DF1XF1-under flour, DF1XF1-upper flour and refined wheat flour belonged to the same category, which was consistent with the results of the principal component analysis. Among the 12 varieties of wheat middling flour, DF1XF1-under and DF1XF1-upper had the best quality, whilst 7M-under, 8M-upper and 2T-under had the worst quality. Comprehensive evaluation of the quality characteristics of wheat middling flour based on principal component analysis and cluster analysis can provide a reference for its high-value utilization in foods.