基于液体样品近红外模型在白酒酒醅分析中的应用  

Application of Near Infrared Model Based on Liquid Sample in the Analysis of the Fermented Grains

作　　者：卢中明 郑敏 刘艳 杜礼泉 谢菲 姚自强 LU Zhongming;ZHENG Min;LIU Yan;DU Liquan;Xie Fei;YAO Ziqiang(Forgood Distillery Co.Ltd.,Sichuan 621000,China)

机构地区：[1]四川绵阳丰谷酒业有限责任公司

出　　处：《酿酒》2019年第6期35-39,共5页Liquor Making

摘　　要：对直接采用白酒固体酒醅建立近红外模型分析酒醅中低含量组分时存在的数据差异进行了系统分析,找到了影响检测数据准确性和灵敏度的关键因素。通过调整白酒酒醅浸取液的处理方式、优化光谱和数据处理方法,显著提高了定标模型的稳定性和准确度,首次建立了基于液体样品近红外定量模型测定酒醅中还原糖、酸度和酒精度的快速、准确的方法,为酒醅组分分析提供了新的思路。结果表明:采用偏最小二乘法(PLS)建立的酒醅浸取液定量模型,能够有效扩大酒醅中还原糖、酸度和酒精度的含量分析范围,显著提高定标模型的稳定性和准确度,模型的决定系数(R^2)分别为0.9686、0.9950、0.9807,预测集标准偏差(RMSEP)分别为0.0427%、0.0791mmol/10g、0.0932%vol。The data differences of low content components in the fermented grains were analyzed systematically by using the near infrared model of the solid fermented grains, and the key factors affecting the accuracy and sensitivity of the data were found. The stability and accuracy of the calibration model were improved by adjusting the treatment method of the fermented grains extract and optimizing the spectrum and data processing method. A fast and accurate method for the determination of reducing sugar, acidity and alcohol content in fermented grains based on near infrared quantitative model of liquid samples was established for the first time. The new idea was provided for the analysis of fermented grains. The results showed that the near infrared quantitative model of the extract of the fermented grains was established by partial least squares method(PLS) could effectively expand the content analysis range of reduced sugar, acidity and alcohol content in fermented grains, and significantly improve the stability and accuracy of the calibration model. The determination coefficients(R^2) of the model were 0.9686, 0.9950 and 0.9807 and the root mean square error of prediction(RMSEP) were 0.0427%, 0.0791 mmol/10 g and 0.0932%vol.

关 键 词：近红外光谱 白酒酒醅 液体样品 定量模型 

分 类 号：TS262.3[轻工技术与工程—发酵工程] F207.3[轻工技术与工程—食品科学与工程]