彭帮柱, 龙明华, 岳田利, 袁亚宏, 赵志华. 用偏最小二乘法及傅立叶变换近红外光谱快速检测白酒酒精度[J]. 农业工程学报, 2007, 23(4): 233-237.
    引用本文: 彭帮柱, 龙明华, 岳田利, 袁亚宏, 赵志华. 用偏最小二乘法及傅立叶变换近红外光谱快速检测白酒酒精度[J]. 农业工程学报, 2007, 23(4): 233-237.
    Peng Bangzhu, Long Minghua, Yue Tianli, Yuan Yahong, Zhao Zhihua. Determination of alcohol concentration in distilled spirit based on PLS-FT-NIRS[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2007, 23(4): 233-237.
    Citation: Peng Bangzhu, Long Minghua, Yue Tianli, Yuan Yahong, Zhao Zhihua. Determination of alcohol concentration in distilled spirit based on PLS-FT-NIRS[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2007, 23(4): 233-237.

    用偏最小二乘法及傅立叶变换近红外光谱快速检测白酒酒精度

    Determination of alcohol concentration in distilled spirit based on PLS-FT-NIRS

    • 摘要: 为了得到白酒工业中酒精度的快速检测技术,将偏最小二乘法与傅立叶变换近红外光谱相结合,通过解析白酒样品的近红外光谱图和对光谱进行不同的预处理,结果表明:用最大最小归一化法预处理光谱,光谱范围选择9747.1~7498.3 cm-1和6102~5446.3 cm-1,采用内部交叉验证建立模型,决定系数(R2)为99.99%,交互验证均方根差(RMSECV)为0.165%,主成分数为4,此条件下建模效果较好;模型进行验证结果表明预测集相关系数(R2)为99.80%,预测标准偏差(RMSEP)为0.264%,模型的预测效果很好,具有较高的精密度和良好的稳定性,能满足生产中白酒酒精度的快速检测要求。

       

      Abstract: In order to measuring alcohol concentration in distilled spirit accurately and quickly, a calibration model was established based on Fourier Transform Near Infrared Spectroscopy with partial least square(PLS-FT-NIRS) by Cross-Validation after the spectra were analyzed and discussed. According to the selected spectra ranges of 9747.1~7498.3 cm-1 and 6102~5446.3 cm-1 and vector normalization processing method, the correlation coefficient(R2) of the model, the root mean square error of cross validation(RMSECV) and the rank are 99.99%, 0.165% and four, respectively. Then the model was tested and evaluated and the result showed the R2 of the test set and the root mean square error of prediction(RMSEP) are 99.80%, 0.264% respectively. The method has been applied to quick determination of alcohol concentration with satisfactory results in the distilled spirit industry.

       

    /

    返回文章
    返回