利用近红外技术对河套原酒入库指标的检测研究

Rapid Detection of Warehousing Indexes of Hetao Base Liquor by Using Near Infrared Technology

研究利用近红外技术建立酒度、总酸、乙酸乙酯、乳酸乙酯、己酸乙酯和丁酸乙酯的快速检测方法。经过反复对模型进行优化、验证和修正,得到R^2线性关系较好,交叉验证均方根差RMSECV较小的最优模型。通过对模型的可靠性、稳健性和实用性评估,表明酒度、总酸、己酸乙酯、乙酸乙酯、乳酸乙酯和丁酸乙酯6项指标模型分析的结果可靠准确,其准确度和精密度分别为100%,100%;99.89%,100%;99.35%,100%;98.33%,100%;96.89%,100%;95.87%,100%。说明模型预测效果好,能满足原酒入库质量等级判定的要求。

In this study, the rapid detection model of alcohol content, total acid, ethyl acetate, ethyl lactate, ethyl caproate and ethyl butyrate in Hetao base liquor has been established by using near infrared technology. Through repeated optimization, verification and correction of the model, the optimum model with good R2 linear relation and good cross validation root-mean-square deviation RMSECV was eventually obtained. Through assessment of the reliability, the robustness and the practicability of the model, it was demonstrated that the analytic results of 6 model indexes including alcohol content, total acids, ethyl caproate, ethyl acetate, ethyl lactate and ethyl butyrate were accurate and reliable with accuracy and precision of 100 %, 100 %; 99.89 %, 100 %; 99.35 %, 100 %; 98.33 %, 100 %; 96.89 %, 100 %; and 95.87 %, 100 %, respectively. The established model had good prediction effects and it could meet the judging requirements of warehousing Hetao base liquor.