Spectroscopy and Spectral Analysis, Volume. 42, Issue 5, 1353(2022)

Research on Deep Sorting Approach Based on Infrared Spectroscopy for High-Value Utilization of Municipal Solid Waste

Bin HU1,1; 2;... Hao FU1,1;, Wen-bin WANG1,1;, Bing ZHANG1,1; 2;, Fan TANG3,3; *;, Shan-wei MA1,1; 2; and Qiang LU1,1; 2; *; |Show fewer author(s)
Author Affiliations
  • 11. School of New Energy, North China Electric Power University, Beijing 102206, China
  • 33. School of Artificial Intelligence, Jilin University, Changchun 130012, China
  • show less
    Figures & Tables(11)
    Original infrared spectra of cellulose (a), vinyl polymers (b), woods (c) and low-value wastes (d)
    SNV pretreated infrared spectra of cellulose (a), vinyl polymers (b), woods (c) and low-value wastes (d)
    MCS pretreated infrared spectra of cellulose (a), vinyl polymers (b), woods (c) and low-value wastes (d)
    DC/Smooth pretreated infrared spectra of cellulose (a), vinyl polymers (b), woods (c) and low-value wastes (d)
    The first (a) and second (b) principal component load analysis spectra
    • Table 1. Residual waste materials

      View table
      View in Article

      Table 1. Residual waste materials

      高值化类别材料名称
      纤维素类打印纸、 草纸、 一次性纸杯、 棉布、 烟头
      烯类聚合物方便面包装盒、 食品包装袋、 快餐包装纸、 奶茶杯、 腈纶标签
      木竹类竹扇、 落叶、 干树枝、 木质铅笔、 一次性筷子
      低值类棒骨、 陶瓷、 贝壳
    • Table 2. The principal component eigenvalues and variance contribution of SNV, MSC and DC/smooth datasets

      View table
      View in Article

      Table 2. The principal component eigenvalues and variance contribution of SNV, MSC and DC/smooth datasets

      预处理方法主成分Z1Z2Z3Z4Z5Z6Z7Z8
      特征值208.35102.0450.5039.2232.7114.996.914.15
      SNV方差贡献率/%43.721.410.68.27.13.11.50.9
      累计方差贡献率/%43.765.175.783.991.094.195.696.5
      特征值861.19429.45207.22161.67107.2047.8336.7724.15
      MSC方差贡献率/%44.622.210.78.45.62.51.91.2
      累计方差贡献率/%44.666.877.585.991.594.095.997.1
      特征值149.7380.3916.7614.346.664.983.282.50
      DC/Smooth方差贡献率/%52.228.05.85.02.31.81.10.9
      累计方差贡献率/%52.280.286.091.093.395.196.297.1
    • Table 3. Comparison of classification model accuracies (based on 72×8 dataset)

      View table
      View in Article

      Table 3. Comparison of classification model accuracies (based on 72×8 dataset)

      预处理方式分类准确率/%均值
      /%
      均方根
      误差/%
      PNNGRNNRDFSVM
      未预处理78.785.785.781.582.93.4
      SNV88.790.188.790.189.40.8
      MSC87.390.187.388.788.41.3
      DC/Smooth95.887.595.897.294.14.4
      均值*90.689.290.692.0--
      均方根误差*4.61.54.64.6--
    • Table 4. Comparison of classification model accuracies (based on 72×5 dataset)

      View table
      View in Article

      Table 4. Comparison of classification model accuracies (based on 72×5 dataset)

      预处理方式分类准确率/%均值
      /%
      均方根
      误差/%
      PNNGRNNRDFSVM
      未预处理77.884.784.787.583.64.1
      SNV95.8100.088.990.393.85.1
      MSC98.688.990.291.792.44.3
      DC/Smooth100.094.495.895.896.52.4
      均值*98.194.491.692.6--
      均方根误差*2.15.63.62.8--
    • Table 5. Comparison of classification accuracies for the four kinds of residual wastes (based on 72×8 DC/Smooth dataset)

      View table
      View in Article

      Table 5. Comparison of classification accuracies for the four kinds of residual wastes (based on 72×8 DC/Smooth dataset)

      垃圾类别DC/Smooth均值
      /%
      均方根
      误差/%
      PNNGRNNRDFSVM
      纤维素类95.090.0100.095.095.04.1
      烯类聚合物95.080.090.095.090.07.1
      木竹类95.090.095.0100.095.04.1
      低值类100.091.6100.0100.097.94.2
    • Table 6. Comparison of classification accuracies for the four kinds of residual wastes (based on 72×5 DC/Smooth dataset)

      View table
      View in Article

      Table 6. Comparison of classification accuracies for the four kinds of residual wastes (based on 72×5 DC/Smooth dataset)

      垃圾类别DC/Smooth均值
      /%
      均方根
      误差/%
      PNNGRNNRDFSVM
      纤维素类100.0100.095.095.097.52.9
      烯类聚合物100.090.095.090.093.84.8
      木竹类100.090.095.0100.096.34.8
      低值类100.0100.0100.0100.0100.00.0
    Tools

    Get Citation

    Copy Citation Text

    Bin HU, Hao FU, Wen-bin WANG, Bing ZHANG, Fan TANG, Shan-wei MA, Qiang LU. Research on Deep Sorting Approach Based on Infrared Spectroscopy for High-Value Utilization of Municipal Solid Waste[J]. Spectroscopy and Spectral Analysis, 2022, 42(5): 1353

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: Research Articles

    Received: Mar. 1, 2021

    Accepted: --

    Published Online: Nov. 10, 2022

    The Author Email:

    DOI:10.3964/j.issn.1000-0593(2022)05-1353-08

    Topics