• Chinese Journal of Lasers
  • Vol. 50, Issue 7, 0708007 (2023)
Mingming Liu1, Desheng Kong1, Yuyan Xiang1, Fengyuan Zhao1, Jing Zhang1, Ruipeng Zhang1, Yamin Gao1, Chenhao Zhi1, Yue Liu1, Maoqiang Xie1、*, Zhi Zhang2, Lu Sun2, Xing Zhao2, Nan Zhang2, and Weiwei Liu2
Author Affiliations
  • 1College of Software, Nankai University, Tianjin 300350, China
  • 2Institute of Modern Optics, Nankai University, Tianjin 300350, China
  • show less
    DOI: 10.3788/CJL221489 Cite this Article Set citation alerts
    Mingming Liu, Desheng Kong, Yuyan Xiang, Fengyuan Zhao, Jing Zhang, Ruipeng Zhang, Yamin Gao, Chenhao Zhi, Yue Liu, Maoqiang Xie, Zhi Zhang, Lu Sun, Xing Zhao, Nan Zhang, Weiwei Liu. Quantitative Analysis of NaCl Aerosols Based on Convolutional Neural Network and Filament‐Induced Fluorescence Spectroscopy[J]. Chinese Journal of Lasers, 2023, 50(7): 0708007 Copy Citation Text show less
    Schematic of filament-induced fluorescence spectrum (FIFS) collection device
    Fig. 1. Schematic of filament-induced fluorescence spectrum (FIFS) collection device
    Filament induced by femtosecond laser in NaCl aerosol
    Fig. 2. Filament induced by femtosecond laser in NaCl aerosol
    Filament-induced fluorescence spectra of NaCl aerosol with five mass concentrations near 589 nm
    Fig. 3. Filament-induced fluorescence spectra of NaCl aerosol with five mass concentrations near 589 nm
    Calibration regression curve of NaCl aerosol’s filament-induced fluorescence spectra
    Fig. 4. Calibration regression curve of NaCl aerosol’s filament-induced fluorescence spectra
    Architecture of one-dimensional CNN for predicting mass concentration of NaCl aerosol
    Fig. 5. Architecture of one-dimensional CNN for predicting mass concentration of NaCl aerosol
    Prediction comparison of each model in full and characteristic spectra. (a)-(d) Full spectrum; (e)-(h) characteristic spectrum
    Fig. 6. Prediction comparison of each model in full and characteristic spectra. (a)-(d) Full spectrum; (e)-(h) characteristic spectrum
    Mass concentration prediction modelR2RMSERPDMAEACC
    MLR0.9910.18510.5810.1440.93
    PLSR0.9680.3266.0180.2590.82
    BPNN0.9850.2308.5720.1670.91
    1D-CNN0.9970.11018.4780.0730.99
    Table 1. Mass concentration prediction results for full spectral data
    Mass concentration prediction modelR2RMSERPDMAEACC
    CR0.8180.7682.5460.6820.40
    PLSR0.9900.19810.0580.1540.90
    BPNN0.9930.15912.8230.1041
    1D-CNN0.9970.11018.7020.0711
    Table 2. Mass concentration prediction results for characteristic spectral data
    Mass concentration prediction modelRMSEMAEACC
    MLR1.1751.1750.43
    PLSR1.1021.0890.18
    BPNN0.6140.6060.64
    1D-CNN0.6870.6760.42
    Table 3. Generalized prediction experimental results for full spectral data
    Mass concentration prediction modelRMSEMAEACC
    CR0.7760.8780.32
    PLSR3.1003.0840.42
    BPNN0.6690.6010.69
    1D-CNN0.3290.3110.87
    Table 4. Generalized prediction experimental results for characteristic spectral data
    Actual mass concentration /(mg·m-3RMSEMAEACC
    Average0.340.310.87
    0.330.310.310
    0.660.100.100.97
    1.320.240.240.97
    1.980.370.340.83
    2.640.240.210.97
    3.300.310.291
    3.960.500.440.93
    4.620.290.261
    5.280.340.311
    6.610.600.571
    Table 5. Generalized prediction experimental results of 1D-CNN in characteristic spectral data
    Mingming Liu, Desheng Kong, Yuyan Xiang, Fengyuan Zhao, Jing Zhang, Ruipeng Zhang, Yamin Gao, Chenhao Zhi, Yue Liu, Maoqiang Xie, Zhi Zhang, Lu Sun, Xing Zhao, Nan Zhang, Weiwei Liu. Quantitative Analysis of NaCl Aerosols Based on Convolutional Neural Network and Filament‐Induced Fluorescence Spectroscopy[J]. Chinese Journal of Lasers, 2023, 50(7): 0708007
    Download Citation