论文部分内容阅读
黄酒中氨基酸态氮的常规检测操作复杂,时效性差。通过研究黄酒的近红外光谱,对比分析了不同的光谱预处理方法:矢量归一化、无光谱处理、一阶导数、多元散射校正、一阶倒数+矢量归一化、一阶倒数+多元散射校正。结果表明,采用矢量归一化预处理方法结合偏最小二乘法所建立的快速检测黄酒中氨基酸态氮的近红外模型效果最佳。其中,R值为0.9603,RMSECV值为0.0414,RPD值为3.58,该方法快速,精度高,可用于黄酒氨基酸态氮的日常快速检测,同样为建立黄酒食品安全质量控制体系提供快速检测手段。
Routine detection of amino acid nitrogen in rice wine is complex and time-consuming. By studying the near infrared spectra of rice wine, different methods of spectrum pretreatment were compared and analyzed: vector normalization, no spectrum processing, first derivative, multivariate scatter correction, first order inverse + vector normalization, first order reciprocal + multiple scattering Correction. The results showed that the NIR model of amino acid nitrogen in rice wine was the best by using vector normalized pretreatment combined with partial least squares method. The R value is 0.9603, the RMSECV value is 0.0414 and the RPD value is 3.58. The method is rapid and accurate and can be used for routine rapid detection of amino acid nitrogen in rice wine. It also provides a rapid detection method for establishing quality control system of rice wine food safety.