Chinese Journal of Lasers, Volume. 49, Issue 8, 0802010(2022)

Elman-Neural-Network Based Prediction of Microsecond Laser Coloring on Stainless Steel

Longda Zhang, Haofa Li, Fengshuo An, Zhiwen Wang*, and Hongyu Zheng
Author Affiliations
  • School of Mechanical Engineering, Shandong University of Technology, Zibo, Shandong 255000, China
  • show less
    Figures & Tables(18)
    Schematic of laser processing system
    Microscopic surface structures of color patch. (a) Yellow patch; (b) orange-red patch; (c) blue patch;(d) amplified blue patch surface
    Elman neural network structure
    Influences of laser frequency and number of laser processing cycles on laser coloring effect
    Influences of laser frequency and scanning speed on laser coloring effect
    Influences of number of laser processing cycles and scanning speed on laser coloring effect
    Flow chart of data processing
    Structures of neural network. (a) net_H; (b) net_S; (c) net_B
    Comparison of actual and predicted hue values
    Comparison of actual and predicted saturation values
    Comparison of actual and predicted brightness values
    Relative error curves of test data
    Predicted laser coloring pattern
    Fabricated laser coloring pattern
    • Table 1. Surface chemical compositions of 304 stainless steel

      View table

      Table 1. Surface chemical compositions of 304 stainless steel

      ElementFeCrNiCMnOSi
      Mass fraction /%68.3118.816.994.020.890.660.32
    • Table 2. Surface chemical compositions of 304 stainless steel after laser irradiation(mass fraction, %)

      View table

      Table 2. Surface chemical compositions of 304 stainless steel after laser irradiation(mass fraction, %)

      ElementFeCrNiOCMnSi
      Yellow66.4118.217.294.522.291.070.21
      Orange-red65.7917.557.145.892.620.880.12
      Blue64.3716.627.837.612.600.740.23
    • Table 3. Training parameters of neural network

      View table

      Table 3. Training parameters of neural network

      Neural networkEpochLearning rateMomentum factorMuMSEAverage error of test values
      net_H4310-50.010.00210.007090.0404
      net_S2810-50.010.0012540.04440.1333
      net_B2510-50.010.00400.009490.0405
    • Table 4. Comparison of actual color and predicted color

      View table

      Table 4. Comparison of actual color and predicted color

      ActualPredicted
      ParameterColorParameterColor
      H:323°,S:47%, B:74%H:315.61°,S:43.59%, B:72.80%
      H:206°,S:54%, B:66%H:219.64°,S:46.31%, B:62.86%
      H:277°,S:35%, B:60%H:264.65°,S:38.58%, B:62.24%
      H:343°,S:11%, B:52%H:352.43°,S:13.14%, B:54.51%
    Tools

    Get Citation

    Copy Citation Text

    Longda Zhang, Haofa Li, Fengshuo An, Zhiwen Wang, Hongyu Zheng. Elman-Neural-Network Based Prediction of Microsecond Laser Coloring on Stainless Steel[J]. Chinese Journal of Lasers, 2022, 49(8): 0802010

    Download Citation

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

    Category: laser manufacturing

    Received: Aug. 18, 2021

    Accepted: Oct. 9, 2021

    Published Online: Mar. 23, 2022

    The Author Email: Zhiwen Wang (wangzhiwen@sdut.edu.cn)

    DOI:10.3788/CJL202249.0802010

    Topics