Spectroscopy and Spectral Analysis, Volume. 42, Issue 5, 1393(2022)

Design of Subwavelength Narrow Band Notch Filter Based on Depth Learning

Shuai-shuai ZHANG1,*... Jun-hua GUO1,1;, Hua-dong LIU1,1;, Ying-li ZHANG1,1;, Xiang-guo XIAO2,2; and Hai-feng LIANG1,1; *; |Show fewer author(s)
Author Affiliations
  • 11. School of Optoelectronics Engineering, Xi'an Technological University, Xi'an 710021, China
  • 22. Xi'an Institute of Applied Optics, Xi'an 710065, China
  • show less
    Figures & Tables(14)
    One via wavelength grating structure diagram
    Neural network structure(a): Forward simulation network; (b): Reverse-design network
    (a) Forward simulation Loss function curve; (b) Inverse design Loss function curve
    Series neural networkR: Expected spectral response; R': Forward simulation prediction spectrum; D: Sample structure of the original training set; D': Reverse design forecast structure. The red frame is the forward simulation network. The Loss function is modified to solve the problem that the network cannot be fitted due to the non-uniqueness of the data. The middle layer is the output of reverse design and the input of forward simulation
    Series network loss function curve
    Red, green and blue are the spectral response curves reported by references, and black curves are
    RCWA numerical simulation curves with inverse design of series networkBlack curve is target spectrum with a reflectivity of 100%; red-green-blue curves are RCWA simulation curves of reverse design with the reflectivity of 98.91%, 99.98% and 99.88% at the peak wavelengthes of 479.5, 551.0 and 607.0 nm, respectively
    • Table 1. Evaluation indexes of network with different hidden layers

      View table
      View in Article

      Table 1. Evaluation indexes of network with different hidden layers

      网络层数消耗时间/s均方误差
      Model_A5无法收敛0.513 905
      Model_B44650.014
      Model_C33820.10
    • Table 2. Evaluation indexes of network with different network structures

      View table
      View in Article

      Table 2. Evaluation indexes of network with different network structures

      网络结构消耗时间/s均方误差
      Model_150, 200, 200, 5091.640.068 427
      Model_250, 200, 200, 200121.890.082 540
      Model_350, 200, 500, 200190.290.031 505
      Model_450, 200, 500, 500373.530.033 413
    • Table 3. Evaluation indexes of network with different Batch sizes

      View table
      View in Article

      Table 3. Evaluation indexes of network with different Batch sizes

      样本数消耗时间/s均方误差
      Batch_size_1323 4830.025 68
      Batch_size_2645740.034 15
      Batch_size_31283810.028 23
      Batch_size_42562820.031 52
      Batch_size_55122680.040 17
    • Table 4. Comparison of structural parameters

      View table
      View in Article

      Table 4. Comparison of structural parameters

      MethodH1
      /nm
      H2/
      nm
      FΛ
      /nm
      N
      1RCWA901500.53602.2
      Network88.441 5150.4760.485 823362.866 62.193 37
      2RCWA701100.73602.1
      Network67.619 93120.378 00.676 11360.900 52.069 56
      3RCWA90900.653602.3
      Network86.095 81104.919 360.620 73350.5692.302 507
    • Table 5. Red, green and blue structural parameters

      View table
      View in Article

      Table 5. Red, green and blue structural parameters

      H1
      /nm
      H2
      /nm
      FΛ
      /nm
      N
      Red (607 nm)60.5104.50.46363.62.04(Si3N4)
      Green(551 nm)59.2105.80.46323.52.04(Si3N4)
      Blue(479.5 nm)58.5106.50.46273.152.04(Si3N4)
    • Table 6. Evaluation Index of correlation coefficient

      View table
      View in Article

      Table 6. Evaluation Index of correlation coefficient

      相关性无相关弱相关中度相关强相关
      rr<0.10.1<r<0.30.3<r<0.50.5<r<1
    • Table 7. Reverse design parameters

      View table
      View in Article

      Table 7. Reverse design parameters

      H1
      /nm
      H2
      /nm
      FΛ/
      nm
      Nr
      Red (607 nm)79.011 7120.7950.584381.6021.9260.685 1
      Green(551 nm)53.246 52107.6090.411326.2872.0520.813 4
      Blue(479.5 nm)45.680 6993.0370.602283.8142.0140.789 6
    Tools

    Get Citation

    Copy Citation Text

    Shuai-shuai ZHANG, Jun-hua GUO, Hua-dong LIU, Ying-li ZHANG, Xiang-guo XIAO, Hai-feng LIANG. Design of Subwavelength Narrow Band Notch Filter Based on Depth Learning[J]. Spectroscopy and Spectral Analysis, 2022, 42(5): 1393

    Download Citation

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

    Category: Research Articles

    Received: Mar. 5, 2021

    Accepted: --

    Published Online: Nov. 10, 2022

    The Author Email: ZHANG Shuai-shuai (863711514@qq.com)

    DOI:10.3964/j.issn.1000-0593(2022)05-1393-07

    Topics