Laser & Optoelectronics Progress, Volume. 61, Issue 4, 0428006(2024)

Aircraft Wake Inversion Based on Bayesian Network in Lidar Detection

Runping Gu, Tong Lu*, and Zhiqiang Wei
Author Affiliations
  • CAUC College of Air Traffic Management, Tianjin 300300, China
  • show less
    Figures & Tables(20)
    Technology roadmap of Bayesian network
    Decision tree of vortex core range based on characteristic parameters
    Bayesian network model
    The parameters of Bayesian network model obtained by training
    Wake vortex influence area and crosswind value area
    Spectral bandwith diagram when wake vortex is detected
    Lidar detection wind field simulation
    Bayesian network inference results of simulation sample. (a) Left vortex nearest detection point; (b) right vortex nearest detection point
    The relative position diagram of lidar and runway
    Eight sets of vortex core inversion results. (a) Example 1; (b) example 2; (c) example 3; (d) example 4; (e) example 5; (f) example 6; (g) example 7; (h) example 8
    • Table 1. Definition and numbering of state set for observation nodes

      View table

      Table 1. Definition and numbering of state set for observation nodes

      VariableState classificationNumbering
      Elevation angle Aaleft1
      aright2
      else3
      Range gate B1,Pleft-11
      Pleft+1,Pright-12
      Pright+1,3
      Pleft4
      Pright5
      Standard deviation ratio R[2.8,∞)1
      [2.67,2.8)2
      [1.42,2.67)3
      [1,1.42)4
      [0,1)5
    • Table 2. Definition and numbering of state set for intermediate nodes

      View table

      Table 2. Definition and numbering of state set for intermediate nodes

      VariableState classificationNumbering
      MAElevation angle of the left vortex core1
      Elevation angle of the right vortex core2
      Else3
      MBLeft side of the left vortex1
      Between the two vortices2
      Right side of the right vortex3
      Closest point to the left vortex core4
      Range gate closest to the left vortex core,which not be the closest point5
      Closest point to the right vortex core6
      Range gate closest to the right vortex core,which not be the closest point7
    • Table 3. Definition and numbering of state set for output nodes

      View table

      Table 3. Definition and numbering of state set for output nodes

      VariableState classificationNumbering
      MRLeft side of the left vortex1
      Between the two vortices2
      Right side of the right vortex3
      Range gate closest to the vortex core(non-nearest point)4
      Coinciding with vortex core5
      Within(0,2.5]m from the vortex core6
      Within(2.5,5]m from the vortex core7
      Within(5,7.5]m from the vortex core8
    • Table 4. Related parameters of A330-300 model and environment

      View table

      Table 4. Related parameters of A330-300 model and environment

      ParameterValue
      Take-off weight m /kg230000
      Span p /m60.30
      Load coefficient Sπ/4
      Take-off speed V /ms-185
      Air density ρ /kgm-31.16
      Gravity acceleration g /ms-29.81
      Normalized eddy dissipation rate ε*0.07
      Normalized Brunt-Väisälä frequency N*0
    • Table 5. Simulation vortex parameters

      View table

      Table 5. Simulation vortex parameters

      ParameterLeft vortexRight vortex
      Position coordinate /m(450,65)(510,65)
      Circulation600600
      Background windMean wind /(m·s-1-1
      Turbulence intensity /(m·s-11
    • Table 6. BN-MSE optimization inference parameters and results

      View table

      Table 6. BN-MSE optimization inference parameters and results

      Calculated parameterLeft vortexRight vortex
      EMS0.137
      BN inference result77
      Standard deviation(4.931,2.217)(4.185,2.285)
      Range gate(25,26)(29,30)
      Radial distance range[452.5,455][512.5,515]
      Estimated angle8.257.25
      Estimated coordinates(448.58,64.80)(508.68,64.80)
      Estimated circulation593591
    • Table 7. Comparison of inference results between the proposed algorithm and traditional algorithms

      View table

      Table 7. Comparison of inference results between the proposed algorithm and traditional algorithms

      ParameterRange methodGradient methodProposed algorithm(without side-wind exclusion)Proposed algorithm(with side-wind exclusion)
      Simulation vortex core position(left/right)(450,65)/(510,65)
      Estimated result(left/right)(445.89,60.68)/506.46,59.94)(445.62,62.63)/(506.20,62.15)(448.27,64.75)/(508.07,64.72)(448.58,64.80)/(508.68,64.80)
      Relative deviation of position(left/right)(6.02 m,7.14 m)(5.54 m,5.88 m)(1.75 m,1.95 m)(1.42 m,1.33 m)
      Relative deviation of circulation(left/right)(-17,-33)(-18,-21)(-13,-10)(-7,-9)
    • Table 8. Lidar scanning parameters at Guangzhou Baiyun international airport

      View table

      Table 8. Lidar scanning parameters at Guangzhou Baiyun international airport

      ParameterLidar 1Lidar 2
      Lidar modelWind3D 6000Wind3D 6000
      LocationNorth end of west runway(19 end)South of the west runway(01 end)
      Scanning strategyPPI,DBS,RHIRHI
      Range resolution /m15015
      Azimuth angle /(°)RHI:16;286
      PPI:0‒359;
      Elevation angle /(°)RHI:0‒1800‒180
      PPI:3;
      DBS:71.38
    • Table 9. Technical specifications of Lidar

      View table

      Table 9. Technical specifications of Lidar

      ParameterSpecification
      Wavelength /nm1550
      Pulse repetition rate /kHz10
      Pulse energy /μJ160
      Pulse width /ns100‒200
      Power consumption /W<300
      Radial velocity measurement range /(ms-1-37.5‒37.5
      Velocity measurement uncertainty /(ms-10.1
      Measurement range /m40‒6000
      Range resolution /m15‒30
    • Table 10. Mean squared error comparison of inversion results between the proposed model and traditional algorithms

      View table

      Table 10. Mean squared error comparison of inversion results between the proposed model and traditional algorithms

      Case numberAircraft typeEstimated valueEMSImprovement compared to traditional method /%
      Position coordinatecirculationGradient methodRange method
      1B763

      (-350.20,64.02)

      (-308.98,62.05)

      (157,160)0.43848.8336.34
      2A359

      (-359.79,48.13)

      (-322.11,61.85)

      (144,129)0.24959.5880.82
      3B744

      (-355.85,47.42)

      (-300.18,48.03)

      (143,110)0.20781.1917.20
      4B77L

      (-356.74,55.29)

      (-323.36,48.67)

      (161,145)0.31665.9574.23
      5B77L

      (-376.17,53.81)

      (-327.44,41.02)

      (125,138)0.42863.5249.13
      6A346

      (-459.39,24.12)

      (-401.66,16.28)

      (181,137)0.54548.3334.16
      7A333

      (-376.59,69.80)

      (-320.53,69.59)

      (176,135)0.25637.1153.24
      8B752

      (-361.41,33.91)

      (-334.93,37.33)

      (111,113)0.21261.0355.04
    Tools

    Get Citation

    Copy Citation Text

    Runping Gu, Tong Lu, Zhiqiang Wei. Aircraft Wake Inversion Based on Bayesian Network in Lidar Detection[J]. Laser & Optoelectronics Progress, 2024, 61(4): 0428006

    Download Citation

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

    Category: Remote Sensing and Sensors

    Received: Jun. 1, 2023

    Accepted: Jun. 19, 2023

    Published Online: Feb. 26, 2024

    The Author Email: Lu Tong (lutong1779@163.com)

    DOI:10.3788/LOP231435

    Topics