基于稻米力学指标主成分分析的质量评价模型

    Model for evaluating paddy rice quality based on principal component analysis of its mechanical indexes

    • 摘要: 为了克服以往评价稻米质量方法的种种弊端,基于稻米力学流变学指标,应用主成分分析的方法,从样本相关矩阵出发,对稻米7个品种的14个主要力学指标进行了分析。根据累计贡献率达到91.1%以上,确定了跟稻米品质相关的3个独立的新指标:主成分PRIN1,硬度因子;主成分PRIN2,弹性因子;主成分PRIN3,黏性因子,并得到各品质的主成分值。以主成分值为指标对稻米进行分类,其结果与根据食味值分类的结果相接近。主成分值与稻米食味值的相关分析表明,主成分1与食味值呈极显著负相关,主成分2与食味值呈显著正相关。通过逐步回归分析,建立了稻米的食味值与主成分之间的回归方程,作为稻米质量评价模型。研究表明,可用该模型进行不同品种稻米的食味值预测和质量评价。

       

      Abstract: In order to overcome every malpractice of the previous methods for appraising paddy rice quality, based on paddy rice mechanics rheological indexes, this paper, embarking from the sample correlation matrix, analyzes fourteen main mechanical characters of seven varieties of paddy rice by the principal component analysis method. According to more than 91.1% of the cumulative variance proportion, three new independent indexes were established,which were related to the quality of paddy rice. Principal component z1, rigidity factor; principal component z2, elasticity factor; principal component z3, viscidity factor, the principal component values of varieties were obtained. Paddy rice could be classified depending on the principal component values, and the results were similar to the classified results depending on taste quality. The results of correlation analysis between the principal component and the taste quality of paddy rice show that, PRIN1 has a closely positive correlation with the taste quality, and PRIN2 has a positive correlation with the taste quality. Regression equation between taste quality and principal component was developed by using Stepwise Regression, which could be used to predict and evaluate quality of paddy rice.

       

    /

    返回文章
    返回