• Spectroscopy and Spectral Analysis
  • Vol. 42, Issue 11, 3368 (2022)
Dong-sheng WANG*, Hai-long WANG1; 2;, Fang ZHANG1; 3; *;, Lin-fang HAN1; 3;, and Yun LI1;
Author Affiliations
  • 1. School of Mechanics and Civil Engineering, China University of Mining & Technology, Beijing 100083, China
  • show less
    DOI: 10.3964/j.issn.1000-0593(2022)11-3368-05 Cite this Article
    Dong-sheng WANG, Hai-long WANG, Fang ZHANG, Lin-fang HAN, Yun LI. Near-Infrared Spectral Characteristics of Sandstone and Inversion of Water Content[J]. Spectroscopy and Spectral Analysis, 2022, 42(11): 3368 Copy Citation Text show less
    Rock sample and its test point location (red dot)
    Fig. 1. Rock sample and its test point location (red dot)
    XDS Smart Probe near infrared spectrum analyzer
    Fig. 2. XDS Smart Probe near infrared spectrum analyzer
    Near-infrared spectra of sandstone with different degrees of saturation(a): Original spectra; (b): First derivative pre-processed spectra
    Fig. 3. Near-infrared spectra of sandstone with different degrees of saturation
    (a): Original spectra; (b): First derivative pre-processed spectra
    Schematic diagram of characteristic variables of near infrared spectrum
    Fig. 4. Schematic diagram of characteristic variables of near infrared spectrum
    SaturationAHFLR
    99%0.865 90.842 60.218 20.259 50.206 4
    83%0.588 90.594 60.285 80.197 70.354 1
    75%0.731 00.661 90.301 60.400 30.186 0
    67%0.588 40.651 80.279 20.259 10.288 0
    59%0.674 80.656 60.318 10.285 70.315 0
    51%0.694 30.446 70.457 40.482 10.317 1
    43%0.201 10.227 30.431 10.547 00.221 5
    35%0.118 60.175 00.536 60.839 70.088 6
    27%0.358 20.284 50.565 80.498 80.445 8
    19%0.169 10.132 10.641 21.00 00.078 3
    11%0.204 00.179 60.552 90.548 10.382 6
    3%0.230 00.160 50.917 00.534 30.880 3
    0%0.304 80.251 10.366 00.387 50.283 6
    Table 1. Normalized characteristic variable
    PeakMICAHFLR
    R1 (1 400 nm)S0.946 00.980 40.642 30.619 40.456 5
    R2 (1 900 nm)S0.960 20.980 40.906 60.620 30.952 4
    Table 2. MIC value caclulated to get the correlation degree between saturation and spectral characteristics
    隐含层神经元个数准确率/%
    训练集测试集
    572.264.5
    686.174.2
    794.474.2
    894.471.0
    993.177.4
    1090.383.9
    1195.871.0
    12100.077.4
    13100.074.2
    14100.071.0
    1597.277.4
    Table 3. Training results of different number of neurons in hidden layer
    类型准确率/%
    0~20%83.3
    21%~40%83.3
    41%~60%83.3
    61%~80%66.7
    81%~100%100
    Training90.3
    Test83.9
    Table 4. Identification results
    Dong-sheng WANG, Hai-long WANG, Fang ZHANG, Lin-fang HAN, Yun LI. Near-Infrared Spectral Characteristics of Sandstone and Inversion of Water Content[J]. Spectroscopy and Spectral Analysis, 2022, 42(11): 3368
    Download Citation