High Power Laser and Particle Beams, Volume. 36, Issue 9, 099002(2024)

Prediction of system-level electric field radiated emission based on ANN reverse model

Luyao Liu1...2,3, Xiao Jin1,4, and Jinliang Cai14 |Show fewer author(s)
Author Affiliations
  • 1Institute of Applied Electronics, CAEP, Mianyang 621999, China
  • 2The 29th Research Institute of China Electronics Technology Group Corporation, Chengdu 610036, China
  • 3Graduate School of China Academy of Engineering Physics, Beijing 100088, China
  • 4National Key Laboratory of Science and Technology on Advanced Laser and High Power Microwave, Mianyang 621900, China
  • show less
    Figures & Tables(13)
    RE102 test setup
    RE102 test data of single equipment
    Equivalence principle of radiated emission (RE)
    Model and simulation of dipole antenna
    Equivalent radiation source of each equipment
    Simulation model
    System RE while single equipment operating
    System RE while all equipments operating
    Three level LM-BP ANN
    • Table 1. Validation dataset

      View table
      View in Article

      Table 1. Validation dataset

      No.f/MHzx1y1z1E1/(dBuV/m)x2y2z2E2/(dBuV/m)x3y3z3E3/(dBuV/m)Etotal/(dBuV/m)
      13000−1004.31000120−1004.31050310−75042.231.939571
      27000−10010.91000120−1003.21050310−75039.440.123183
      310000−1009.21000120−10010.71050310−75045.253.036382
      413000−10030.51000120−10011.51050310−75045.360.507379
      516000−10019.91000120−10011.61050310−75048.364.899009
      620000−10020.91000120−10017.31050310−75057.277.367965
      7300100−2004.3900200−2004.31000400−60042.235.062737
      8700100−20010.9900200−2003.21000400−60039.443.601726
      91000100−2009.2900200−20010.71000400−60045.256.07834
      101300100−20030.5900200−20011.51000400−60045.362.877408
      2530−2000−4004.31200150−1504.31200200−80042.232.92811
      2670−2000−40010.91200150−1503.21200200−80039.440.90428
      27100−2000−4009.21200150−15010.71200200−80045.254.0078
      28130−2000−40030.51200150−15011.51200200−80045.361.66204
      29160−2000−40019.91200150−15011.61200200−80048.365.79797
      30200−2000−40020.91200150−15017.31200200−80057.277.72475
    • Table 2. Validation error and relative coefficient of different number of training dataset

      View table
      View in Article

      Table 2. Validation error and relative coefficient of different number of training dataset

      number of training datasetvalidation error/%relative coefficient
      307.70080.85303
      1000.87130.87193
      2000.41590.95546
      5000.00590.99809
      10003.4247E-50.99977
      100002.2718E-50.99991
    • Table 3. Alternative neural network

      View table
      View in Article

      Table 3. Alternative neural network

      No.alternative neural networkvalidation error/%
      1x1,c=fann,1(y, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13)0.149589782
      2x2,c=fann,2(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13)0.000017400
      3x3,c=fann,3(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13)0.096499053
      4x4,c=fann,4(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13)0.003367234
      5x5,c=fann,5(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13)3.945121882
      6x6,c=fann,6(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13)0.000000768
      7x7,c=fann,7(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13)0.20383497
      8x8,c=fann,8(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13)0.000812276
      9x9,c=fann,9(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13)26.19090216
      10x10,c=fann,10(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13)8.284504188
      11x11,c=fann,11(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13)0.061618247
      12x12,c=fann,12(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13)0.430195851
      13x13,c=fann,13(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13)10.20954216
    • Table 4. Comparison of prediction results

      View table
      View in Article

      Table 4. Comparison of prediction results

      No.Etotal/(dBμV·m−1)
      CST simulationtraditionalLM-BP ANNANN reverse model(conjugate gradient method)ANN reverse model(regular-falsi method)
      131.9445.6430.3431.04
      240.1252.9138.1139.00
      353.0456.4152.9151.55
      460.5148.0360.3358.81
      564.9058.9865.9863.08
      677.3760.8676.6375.20
      735.0654.0426.0234.08
      843.6055.3836.4342.38
      956.0860.1249.3154.51
      1062.8856.2865.8961.11
      1167.5860.8466.3665.68
      1280.0963.8574.1777.84
      1333.3760.0033.6832.44
      1440.8862.7841.7239.73
      1553.7460.1353.2152.23
      1660.8761.5460.1559.17
      1765.8465.6266.1363.99
      1878.8668.2178.3676.65
      1932.8047.9127.0531.88
      2041.0253.7037.3339.87
      2153.8055.9148.4952.29
      2261.0952.6959.8859.38
      2365.6759.8560.6163.83
      2478.1464.5665.7575.95
      2532.9358.0231.7732.00
      2640.9057.6438.7639.76
      2754.0156.9053.7052.49
      2861.6660.1260.5259.93
      2965.8061.4367.0463.95
      3077.7266.6377.6375.55
    Tools

    Get Citation

    Copy Citation Text

    Luyao Liu, Xiao Jin, Jinliang Cai. Prediction of system-level electric field radiated emission based on ANN reverse model[J]. High Power Laser and Particle Beams, 2024, 36(9): 099002

    Download Citation

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

    Category: Advanced Interdisciplinary Science

    Received: May. 24, 2024

    Accepted: Jul. 22, 2024

    Published Online: Oct. 15, 2024

    The Author Email:

    DOI:10.11884/HPLPB202436.240177

    Topics