《食品工业科技》F5000论文展播：基于主成分分析和聚类分析综合评价蒸谷米的品质特性（2022年）

Objectives:The parboiled rice has more advantages including easy digestion and absorption, compact grain structure, high percentage of brown rice, low percentage of broken rice, high percentage of rice and storage. At present, there are few reports on the systematic analysis and evaluation of quality characters and gelatinization properties of parboiled rice prepared from different rice varieties. Twelve different rice raw materials were selected and prepared into parboiled rice. The quality characteristics and gelatinization properties of parboiled rice were systematically analyzed to explore the changes in the quality characteristics of parboiled rice derived from different rice varieties, and screening the most suitable rice varieties for preparing parboiled rice, also selecting the key proxies affecting the quality of parboiled rice.

Methods:Descriptive statistics, correlation analysis, gelatinization properties analysis, principal component analysis (PCA) and cluster analysis (CA) were used to explore the quality characteristics of parboiled rice and establish its evaluation system. The first thing was to use 12 varieties parboiled rice for quality (milling, appearance, nutrition, taste and gelatinization) analysis. The quality of parboiled rice was comprehensively evaluated by PCA and CA. The special rice varieties suitable for preparing parboiled rice and the key indexes affecting the quality were screened out through the evaluation model. 

Results:Descriptive statistics showed that the measured values of each index of the selected parboiled rice materials were within the acceptable range, and their quality characteristics were quite different, which had a certain degree of generality and representativeness. Comprehensive correlation analysis and gelatinization properties analysis found that there was a certain correlation between the quality (milling, appearance, nutrition, taste and gelatinization) of parboiled rice, and the information presented by the measured values of each index appeared overlapping phenomenon. Therefore, PCA and CA could be applied to simplify the data structure of the quality index of parboiled rice via dimensionality reduction, improving the valuation precision of different varieties of parboiled rice quality characteristics CA found when the Euclidean distance 2.5, parboiled rice quality indexes can be divided into 5 major categories, and combined with principal component load analysis, 8 quality indexes including the crack ratio, yellow rice, appearance, protein, the breakdown, the peak viscosity, the final viscosity and setback first were sufficient to evaluate the quality of parboiled rice. As the Euclidean distance was 5, the 12 different parboiled rice varieties could be divided into 4 categories. Combined with the factor score and comprehensive score results of PCA, the parboiled rice produced by ‘Yuzhenxiang’ had the highest comprehensive score, followed by ‘Guifeng you 785’ and ‘Jinmazhan’, which were grouped into the same category with ‘Yuzhenxiang’. Therefore, there was closely relevent between the parboiled rice quality of the three rice varieties. The results showed that ‘Yuzhenxiang’, ‘Guifengyou 785’ and ‘Jinmazhan’ all belonged to the category of better quality, while ‘Dalixiang’ had the worst quality, followed by ‘Yuefeng B’.

Conclusion:Through the comprehensive analysis of the quality characteristics and gelatinization properties of parboiled rice prepared from 12 different rice varieties, the differences between the quality indicators and the changes of differences in the quality traits of parboiled rice can be further judged. Meanwhile, PCA and CA can effectively provide theoretical basis and technical guidance for the screening of suitable rice varieties and  determining key indexes affecting the quality of parboiled rice.