• Spectroscopy and Spectral Analysis
  • Vol. 42, Issue 7, 2025 (2022)
Xiao-qin LIAN*, Qun CHEN1; 2;, Shen-miao TANG1; 2;, Jing-zhu WU1; 2;, Ye-lan WU1; 2;, and Chao GAO1; 2;
Author Affiliations
  • 1. School of Artificial Intelligence, Beijing Technology and Business University, Beijing 100048, China
  • show less
    DOI: 10.3964/j.issn.1000-0593(2022)07-2025-08 Cite this Article
    Xiao-qin LIAN, Qun CHEN, Shen-miao TANG, Jing-zhu WU, Ye-lan WU, Chao GAO. Quantitative Analysis Method of Key Nutrients in Lanzhou Lily Based on NIR and SOM-RBF[J]. Spectroscopy and Spectral Analysis, 2022, 42(7): 2025 Copy Citation Text show less
    Vertex70 Fourier transform infrared spectrometer
    Fig. 1. Vertex70 Fourier transform infrared spectrometer
    NIR spectra of the SG+Detrend method
    Fig. 2. NIR spectra of the SG+Detrend method
    NIR spectra of the Detrend method
    Fig. 3. NIR spectra of the Detrend method
    SG+Detrend_SPA_PLSR method for the prediction results of protein
    Fig. 4. SG+Detrend_SPA_PLSR method for the prediction results of protein
    Detrend_SPA_PLSR method for the prediction results of polysaccharide
    Fig. 5. Detrend_SPA_PLSR method for the prediction results of polysaccharide
    SOM neural network structure of protein and polysaccharide
    Fig. 6. SOM neural network structure of protein and polysaccharide
    RBF network topology of protein and polysaccharide
    Fig. 7. RBF network topology of protein and polysaccharide
    SOM-RBF prediction algorithm flow chart for protein and polysaccharide[10]
    Fig. 8. SOM-RBF prediction algorithm flow chart for protein and polysaccharide[10]
    SOM-RBF network prediction results of protein
    Fig. 9. SOM-RBF network prediction results of protein
    SOM-RBF network prediction results of polysaccharide
    Fig. 10. SOM-RBF network prediction results of polysaccharide
    营养
    物质
    样本集样本
    个数
    最小值
    g/100 g
    最大值
    g/100 g
    平均值
    g/100 g
    标准
    偏差
    蛋白质建模集454.9314.209.421.86
    预测集144.9312.109.611.82
    多糖建模集4516.7832.5721.863.16
    预测集1417.3529.6521.643.57
    Table 1. Basic statistics of Lanzhou lily samples
    处理参数预处理方法RcRMSECRvRMSECVRpRMSEP
    蛋白质None0.824 80.910 70.706 01.157 90.771 81.424 4
    SG0.824 10.912 30.705 91.158 00.771 71.424 3
    Normalize0.895 30.719 60.835 50.891 80.846 81.130 5
    SNV0.864 20.850 70.748 41.134 60.848 60.992 5
    MSC0.801 80.920 80.666 51.163 60.759 01.212 4
    Detrend0.849 90.912 70.778 21.094 20.822 41.256 1
    OSC0.996 00.166 90.870 50.917 00.346 21.890 6
    SG+1D0.965 00.461 70.672 21.307 80.778 11.181 2
    SG+Normalize0.808 61.046 30.693 21.294 70.796 91.221 9
    SG+SNV0.749 41.067 10.599 71.306 80.814 21.138 1
    SG+Detrend0.915 30.699 40.827 50.989 00.870 11.081 1
    Table 2. PLSR modeling results of Lanzhou lily protein using the spectra pretreated by different methods
    处理参数预处理方法RcRMSECRvRMSECVRpRMSEP
    多糖None0.888 91.446 80.770 42.038 60.777 02.317 5
    SG0.887 41.455 80.771 12.036 90.777 02.316 8
    Normalize0.877 81.512 50.736 22.181 80.740 72.438 6
    SNV0.926 71.186 70.826 11.787 50.949 61.623 2
    MSC0.941 00.919 80.833 71.511 30.901 81.648 8
    Detrend0.966 70.697 30.903 11.171 70.921 61.692 1
    OSC0.999 30.117 00.966 60.809 80.835 92.131 1
    SG+1D0.953 60.941 70.518 32.696 60.452 21.241 4
    SG+Normalize0.917 91.225 20.838 81.690 20.632 12.819 4
    SG+SNV0.946 50.877 30.850 21.438 70.914 31.611 4
    SG+Detrend0.933 21.135 00.846 61.688 10.909 81.914 7
    Table 3. PLSR modeling results of Lanzhou lily polysaccharide using the spectra pretreated by different methods
    处理参数预处理特征提取法特征波长数RcRMSECRvRMSECVRpRMSEP
    蛋白质SG+DetrendNone8310.915 30.699 40.825 70.989 00.870 11.081 1
    SG+DetrendCARS40.882 60.774 80.846 70.880 80.867 21.144 9
    SG+DetrendSPA20.884 50.722 80.853 80.807 60.810 61.195 3
    SG+DetrendPCA30.787 51.084 10.741 51.180 40.816 41.218 3
    多糖DetrendNone8310.966 70.697 30.903 11.171 70.921 61.692 1
    DetrendCARS130.929 11.000 10.874 41.329 30.885 11.668 1
    DetrendSPA140.967 80.632 70.920 30.988 60.810 92.094 6
    DetrendPCA30.737 21.896 20.501 52.487 70.225 34.402 6
    Table 4. Comparison of PLSR modeling results based on CARS, SPA and PCA methods
    营养物质预测方法RpRMSEP
    蛋白质PLSR0.810 61.195 3
    SOM-RBF0.866 61.038 5
    多糖PLSR0.810 92.094 6
    SOM-RBF0.868 11.799 4
    Table 5. Comparison of modeling results between PLSR and SOM-RBF prediction models
    Xiao-qin LIAN, Qun CHEN, Shen-miao TANG, Jing-zhu WU, Ye-lan WU, Chao GAO. Quantitative Analysis Method of Key Nutrients in Lanzhou Lily Based on NIR and SOM-RBF[J]. Spectroscopy and Spectral Analysis, 2022, 42(7): 2025
    Download Citation