摘要
采用皂化返滴定法测定了酒醅中总酯的含量,并利用近红外光谱技术建立了酒醅总酯的快速检测方法。通过解析不同酒醅样品的近红外光图谱,采用MF和NCL光谱前处理方法对光谱进行了平滑和去噪处理,用偏最小二乘(PLS)算法建立了酒醅总酯的近红外模型。模型的决定系数R^2为97.15%,预测标准偏差(RMSEP)为0.07%,表明建立的模型可靠,预测效果好,能够满足白酒生产中酒醅总酯的快速检测要求。
The total ester in fermented grains was determined by saponification back titration, and near infrared spectroscopy was developed as a new method to rapid determine the total ester in fermented grains. After being smoothed and processed by MF and NCL spectral pretreatment methods, the near infrared spectra of total ester from different fermented grains were analysed and we established a total ester detection model by using the partial least squares （PLS） algorithm. Results show that the correlation coefficient R^2 and the root mean square errors of prediction （RMSEP） for the model are 97.15% and 0.07% respectively, which suggest that the model is reliable with great predictability and can satisfy the requirement of quick detection of total ester during liquor production.
出处
《酿酒》
CAS
2017年第3期83-85,共3页
Liquor Making
关键词
酒醅
总酯
近红外光谱技术
fermented grains
total ester
Near infrared spectroscopy
