近红外光谱法快速检测白酒基酒中甲醇含量
|
摘要
使用近红外光谱类型检测技术对白酒基酒当中的甲醇物质含量进行了检验,在完成了光谱信息数据采集之后对数据进行了有效处理,运用偏最小二乘法建立起甲醇物质定量分析模型。在对模型进行了优化处理之后,确定最佳效果的预处理方式是将一阶导数方程和矢量均归一化,分析得主要成分数数值为2,最佳波段的谱值在6102到5450厘米之间。通过相应处理方式使待处理样品的真实数值和预测数值的具体决定系数数值为0.955,将标准偏差数值校正为1.30。在完成了分析之后,对相应的模型进行科学验证,最终验证集的决定系数数值调整为0.950,预测得出的标准偏差数值为1.83。最终的研究结果表明:该试验当中确立起的甲醇物质分析模型基础和常见类型的气象色谱分析法有一致的结构,在试验中能保证试验分析有良好的稳定性以及精密度,这样也就能使得在线类型快速检测技术能白酒当中酒基是实际含量检测中发挥良好作用。
The content of methanol in liquor-based liquor was tested by near infrared spectral type detection technique, the data was processed effectively after spectral information data collection, and the quantitative analysis model of methanol was established by partial least square method. After optimizing the model, the pretreatment method of determining the best effect is to normalized the first derivative equation and vector, the main component number is 2, and the best band's spectral value is between 6102 and 5450 centimeters. The value of the real and predictive values of the sample to be processed is 0.955, and the calibration deviation is corrected to 1.30 by the corresponding processing method. The final results show that the established methanol substance analysis model and common types of meteorological chromatographic analysis have a consistent structure in the experiment, which can ensure good stability and precision in the test. This will enable the online type of rapid detection technology can play a good role in liquor based on the actual content detection.
The content of methanol in liquor-based liquor was tested by near infrared spectral type detection technique, the data was processed effectively after spectral information data collection, and the quantitative analysis model of methanol was established by partial least square method. After optimizing the model, the pretreatment method of determining the best effect is to normalized the first derivative equation and vector, the main component number is 2, and the best band's spectral value is between 6102 and 5450 centimeters. The value of the real and predictive values of the sample to be processed is 0.955, and the calibration deviation is corrected to 1.30 by the corresponding processing method. The final results show that the established methanol substance analysis model and common types of meteorological chromatographic analysis have a consistent structure in the experiment, which can ensure good stability and precision in the test. This will enable the online type of rapid detection technology can play a good role in liquor based on the actual content detection.
引文
[1]刘建学,郭玉姗,李璇,等.近红外光谱法快速检测白酒基酒中甲醇含量[J].河南科技大学学报(自然科学版),2018(5).
[2]周玮婧,江小明.顶空气相色谱法测定葡萄酒中甲醇及市售葡萄酒质量调查[J].酿酒科技,2016(11):116-119.
[3]熊建军,杨坤,张建炀,等.凯氏定氮-气相色谱法测定蓝莓利口酒中微量甲醇的含量[J].现代食品,2018(11).
[4]黄志鸿.基于近红外、中红外和拉曼光谱法甲醇柴油品质检测研究[D].华东交通大学,2016.
[5]欧阳爱国,唐天义,周鑫,等.最小二乘支持向量机结合中红外光谱测定甲醇柴油甲醇含量[J].发光学报,2016,37(10):1253-1258.
[2]周玮婧,江小明.顶空气相色谱法测定葡萄酒中甲醇及市售葡萄酒质量调查[J].酿酒科技,2016(11):116-119.
[3]熊建军,杨坤,张建炀,等.凯氏定氮-气相色谱法测定蓝莓利口酒中微量甲醇的含量[J].现代食品,2018(11).
[4]黄志鸿.基于近红外、中红外和拉曼光谱法甲醇柴油品质检测研究[D].华东交通大学,2016.
[5]欧阳爱国,唐天义,周鑫,等.最小二乘支持向量机结合中红外光谱测定甲醇柴油甲醇含量[J].发光学报,2016,37(10):1253-1258.