桂林米粉入味性质评价模型及预测模型的建立

Establishment of Evaluation Model and Prediction Model for Taste Properties of Guilin Rice Noodles

摘要:

测定了桂林米粉的吸盐率、吸油率、色差值、孔隙率、卤水吸收率、卤水风味相似度、感官评价等 7 个入味性评价指标。其中各组样品吸盐率为 1.05% - 1.80%，吸油率 0.4% - 1.0%，卤水吸收率 1.69% - 3.93%，孔隙率 13.56% - 34.56%，色差值 20.42 - 31.61，风味物质相似度 25.49 - 32.59，感官评价结果为 41.71 - 49.75 分，对这些指标进行主成分分析，前两个主成分累计贡献率达 89.08%，确定主成分与米粉入味性质指标的线性关系表达式，从而建立桂林米粉入味性质评价模型。再测定桂林米粉的断条率、水分含量、吐浆值、糊汤度、碘蓝值、复水率及质构指标等品质评价指标，以桂林米粉入味评价综合得分为因变量，桂林米粉的品质评价指标为自变量，利用逐步回归分析对影响桂林米粉入味性质的桂林米粉品质指标进行筛选并建立起了桂林米粉入味性质预测模型，预测模型以硬度、断条率、弹性、咀嚼性为参考指标。两组模型分别从消费者和生产加工角度出发，为桂林米粉评价体系完善及品质提升提供参考。

Abstract:

Guilin rice noodle were tested for seven taste evaluation indexes, including salt absorption rate, oil absorption rate, color difference value, porosity, brine absorption rate, brine flavor similarity, and sensory evaluation. The salt absorption rate of each group of samples is 1.05% -1.80%, oil absorption rate is 0.4% -1.0%, brine absorption rate is 1.69% -3.93%, porosity is 13.56% -34.56%, color difference value is 20.42-31.61, flavor substance similarity is 25.49-32.59, and sensory evaluation result is 41.71-49.75 points. Through principal component analysis of these indicators, the cumulative contribution rate of the first two principal components is 89.08%, and the linear relationship expression between the principal components and Rice noodles flavor property indicators is determined, so as to establish the evaluation model of Guilin rice noodle taste properties. Then measure the quality evaluation indexes of Guilin rice noodle, such as broken bar rate, moisture content, spit value, paste degree, iodine blue value, rehydration rate and texture index. With the comprehensive score of Guilin rice noodle as the dependent variable, and the quality evaluation index of Guilin rice noodle as the independent variable, use stepwise regression analysis to screen the quality indexes of Guilin rice noodle that affect the taste properties of Guilin rice noodle and establish a prediction model of Guilin rice noodle taste properties. The prediction model takes hardness, broken bar rate, elasticity, and chewiness as the reference indexes. The two sets of models provide reference for the improvement of Guilin rice noodle evaluation system and quality improvement from the perspective of consumers and production and processing.