基于主成分分析和聚类分析的食盐综合评价

Comprehensive Evaluation of Salt Based on Principal Component Analysis and Cluster Analysis

为了解我国食盐产品的品质现状,文章对28种市售食盐的成分和滋味进行解析,通过相关性分析、主成分分析、建立综合评价模型和聚类分析方法对食盐样品进行评价和分类。结果表明,井矿盐氯化钠含量较高,最高可达99.72%;海盐氯化钠含量较低,但钙、镁、钾等无机元素丰富。食盐样品5种滋味指标间呈现显著性差异(P<0.05);在10种品质评价指标中提取了4个主成分,累计方差贡献率达79.041%;建立综合评价模型,发现海盐得分最高,井矿盐次之,湖盐得分最低;在欧氏距离为10时,可将食盐样品分为4类,符合样品实际情况。综上,主成分分析和聚类分析效果良好,将其应用于品质评价领域可为食盐的质量控制提供科学依据。

In order to understand the quality status of salt products in China,in this paper,the components and taste of 28 kinds of commercially available salt are analyzed,and the salt samples are evaluated and classified by correlation analysis,principal component analysis,establishing comprehensive evaluation model and cluster analysis.The results show that the content of sodium chloride in well and rock salt is high,up to 99.72%.The content of sodium chloride in sea salt is low,but it is rich in inorganic elements such as calcium,magnesium and potassium.There are significant differences in five taste indexes of salt samples(P<0.05).Four principal components are extracted in ten quality evaluation indexes,and the cumulative variance contribution rate is 79.041%.The comprehensive evaluation model is established,and it is found that the score of sea salt is the highest,followed by well and rock salt,and the score of lake salt is the lowest.When the Euclidean distance is 10,the salt samples could be divided into four categories,which confirms to the actual situation.In conclusion,the effects of principal component analysis and cluster analysis are good,and their application in the field of quality evaluation can provide scientific basis for the quality control of salt.