Acta Photonica Sinica, Volume. 54, Issue 3, 0312004(2025)

Application of Improved Whale Optimization Algorithm in Particle Size Distribution Inversion for Forward Laser Scattering Particle Measurement Technology

Huiling LIU1, Xingxing HAN1、*, Bei ZHAO1,2, Bing GAO3, and Jiajie WANG3
Author Affiliations
  • 1School of Physics and New Energy,Xi'an Jiaotong University City College,Xi'an 710018,China
  • 2Engineering Research Center of Photovoltaic Technologies and Systems,Universities of Shaanxi Province,Xi'an 710018,China
  • 3School of Physics,Xidian University,Xi'an 710071,China
  • show less
    Figures & Tables(18)
    Light path diagram of small angle forward scattering method
    Diagram and image of a multichannel ring photodetector receiving scattered light
    Flowchart of improved whale optimization algorithm
    Comparisons of inversion results and theoretical results for unimodal distributions
    The inversion results of bimodal distribution
    The inversion results of trimodal distribution
    Comparison of theoretical and measured light energy distribution of 80 μm standard particles
    Inversion results of 2 drops of 80 μm standard particles by IWOA
    Comparison of theoretical and measured light energy distribution of mixed particles
    Inversion results of mixed standard particles by IWOA
    • Table 1. Inversion results of unimodal distribution for different distributions

      View table
      View in Article

      Table 1. Inversion results of unimodal distribution for different distributions

      Distribution functionPreset value(μσInversion result(μσStandard deviation(μσRRMSE/%Time/s
      Normal distribution(45.00,10.00)(45.00,10.00)(0,0)01.06
      RR distribution(45.00,10.00)(45.00,10.00)(0,0)01.70
      JSB distribution(45.00,10.00)(45.00,10.00)(0,0)01.28
    • Table 2. Inversion results of bimodal distribution for each distribution type

      View table
      View in Article

      Table 2. Inversion results of bimodal distribution for each distribution type

      Distribution functionPreset value(AratμσInversion result(AratμσStandard deviation(μσRRMSE/%Time/s
      Normal distribution

      (0.500,30.00,6.00;

      0.500,70.00,6.00)

      (0.500,30.00,6.00;

      0.500,70.00,6.00)

      (0,0;

      0,0)

      01.33

      RR

      distribution

      (0.500,30.00,6.00;

      0.500,70.00,6.00)

      (0.500,30.00,6.00;

      0.500,70.00,6.00)

      (0,0;

      0,0)

      02.17

      JSB

      distribution

      (0.500,30.00,6.00;

      0.500,70.00,6.00)

      (0.500,30.00,6.00;

      0.500,70.00,6.00)

      (0,0;

      0,0)

      01.83
    • Table 3. Inversion results of trimodal distribution for each distribution type

      View table
      View in Article

      Table 3. Inversion results of trimodal distribution for each distribution type

      Distribution functionPreset value(AratμσInversion result(AratμσStandard deviation(μσRRMSE/%Time/s
      Normal distribution

      (0.300,15.00,5.00;

      0.400,50.00,5.00;

      0.300,90.00,5.00)

      (0.299,15.00,5.00;

      0.398,50.00,4.97;

      0.303,90.20,5.12)

      (0.014,0.016;

      0.118,0.282;

      0.666,1.062)

      1.674.66

      RR

      distribution

      (0.300,15.00,5.00;

      0.400,50.00,5.00;

      0.300,90.00,5.00)

      (0.297,14.98,5.03;

      0.361,49.85,5.09;

      0.342,87.43,4.40)

      (0.081,0.096;

      1.130,0.410;

      4.710,1.203)

      3.0710.18

      JSB

      distribution

      (0.300,15.00,5.00;

      0.400,50.00,5.00;

      0.300,90.00,5.00)

      (0.292,15.00,5.00;

      0.370,49.53,5.19;

      0.338,90.54,4.47)

      (0.007,0.020;

      0.718,0.374;

      2.961,1.272)

      3.517.60
    • Table 4. The inversion results of normal distribution with random noise

      View table
      View in Article

      Table 4. The inversion results of normal distribution with random noise

      Preset value(AratμσRandom noiseInversion result(AratμσStandard deviation(μσRRMSE/%Time/s

      Unimodal distribution

      (1.000,45.00,10.00)

      1%

      3%

      5%

      (1.000,44.98,9.99)

      (1.000,44.93,9.96)

      (1.000,44.90,9.92)

      (0.046,0.057)

      (0.182,0.157)

      (0.262,0.269)

      0.15

      0.58

      0.95

      4.32

      2.86

      2.90

      Bimodal distribution1%

      (0.500,30.01,6.01;

      0.500,70.03;6.02)

      (0.052,0.029;

      0.206,0.312)

      0.383.71

      (0.500,30.00,6.00;

      0.500,70.00,6.00)

      3%

      (0.501,30.02,6.01;

      0.499,70.17,6.28)

      (0.192,0.139;

      0.596,0.902)

      3.143.68
      5%

      (0.502,30.07,6.02;

      0.498,70.41,6.15)

      (0.211,0.155;

      0.950,1.878)

      3.793.71

      Trimodal distribution(0.300,15.00,5.00;

      0.400,50.00,5.00;

      0.300,90.00,5.00)

      1%

      3%

      5%

      (0.296,14.99,4.99;

      0.395,49.98,4.98;

      0.309,90.41,5.54)

      (0.295,14.98,4.99;

      0.395,50.06,5.23;

      0.310,90.27,5.46)

      (0.295,15.03,5.06;

      0.395,50.24,5.00;

      0.310,91.06,4.37)

      (0.047,0.048;

      0.253,0.325;

      1.528,1.134)

      (0.118,0.134;

      0.639,0.929;

      1.904,1.670)

      (0.189,0.191;

      1.015,1.431;

      4.237,1.645)

      4.49

      4.78

      11.77

      4.64

      6.16

      6.22

    • Table 5. Comparison of inversion results of four algorithms

      View table
      View in Article

      Table 5. Comparison of inversion results of four algorithms

      Preset value(AratμσRandom noiseProjection RRMSE/%Chahine RRMSE/%ABC RRMSE/%IWOA RRMSE/%
      (1,30,2)

      0%

      5%

      4.96

      16.49

      3.49

      14.13

      0.02

      1.48

      0

      0.33

      (1,30,4)

      0%

      5%

      3.54

      16.65

      1.16

      14.31

      2.27×10-8

      3.65

      0

      0.45

      (1,30,6)

      0%

      5%

      4.55

      18.46

      0.20

      12.13

      1.72×10-5

      2.41

      0

      0.85

      (0.5,30,2;0.5,60,2)

      0%

      5%

      5.75

      23.19

      2.70

      25.56

      1.53

      10.34

      0

      10.77

      (0.5,30,4;0.5,60,4)

      0%

      5%

      2.47

      24.02

      5.06

      25.17

      0.07

      5.75

      0

      2.75

      (0.5,30,6;0.5,60,6)

      0%

      5%

      5.49

      26.97

      6.36

      24.36

      1.23

      18.22

      0

      5.69

    • Table 6. Parameter information for standard particles

      View table
      View in Article

      Table 6. Parameter information for standard particles

      Standard material numberStandard particle sizeCoefficient of variation
      GBW(E)120030(80 μm)Number mean diameter:74.9 μmNumber median diameter:74.9 μm4.5%
      Volume mean diameter:76.6 μmVolume median diameter:76.6 μm
      GBW(E)120030(160 μm)Number mean diameter:164.4 μmNumber median diameter:163.7 μm4.8%
      Volume mean diameter:165.6 μmVolume median diameter:165.6 μm
    • Table 7. Inversion results of 2 drops of 80 μm standard particles by IWOA

      View table
      View in Article

      Table 7. Inversion results of 2 drops of 80 μm standard particles by IWOA

      Theoretical characteristic valueMaximum valueRelative errorTime/s
      IWOA74.9 μm80.1 μm0.0696.58
    • Table 8. Inversion results of mixed standard particles by IWOA

      View table
      View in Article

      Table 8. Inversion results of mixed standard particles by IWOA

      Theoretical characteristic valueMaximum valueRelative errorTime/s
      IWOA74.9 μm164.4 μm83 μm146 μm0.108,0.1126.91
    Tools

    Get Citation

    Copy Citation Text

    Huiling LIU, Xingxing HAN, Bei ZHAO, Bing GAO, Jiajie WANG. Application of Improved Whale Optimization Algorithm in Particle Size Distribution Inversion for Forward Laser Scattering Particle Measurement Technology[J]. Acta Photonica Sinica, 2025, 54(3): 0312004

    Download Citation

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

    Category: Instrumentation, Measurement and Metrology

    Received: Oct. 9, 2024

    Accepted: Dec. 25, 2024

    Published Online: Apr. 22, 2025

    The Author Email: Xingxing HAN (hxx_xd@163.com)

    DOI:10.3788/gzxb20255403.0312004

    Topics