obsidian_backup/多体+耦合求解器/SD ED AD 流程及数据交互/流程.md

fast_advance_states()

• ed_updatestates

• SolveOption2a_Inp2BD

  • ed_calc_output 更新数据

• BD updatestates

• SolveOption2b_Inp2IfW Ifw根据ED结果更新

• ifw updatestates

• SolveOption2c_Inp2AD_SrvD

  • ifw calcoutput
  • AD_InputSolve_IfW
  • SrvD_InputSolve
    • ! from Ifw
    • WindDir
    • HorWindV
    • LidSpeed
    • MsrPositionsX
    • MsrPositionsY
    • MsrPositionsZ
    • ! from ED
    • YawAngle
    • Yaw
    • YawRate
    • LSS_Spd
    • HSS_Spd
    • RotSpeed
    • RootMxc
    • RootMyc
    • YawBrTAxp
    • YawBrTAyp
    • LSSTipPxa
    • LSSTipMxa
    • LSSTipMya
    • LSSTipMza
    • LSSTipMys
    • LSSTipMzs
    • YawBrMyn
    • YawBrMzn
    • NcIMURAxs
    • NcIMURAys
    • NcIMURAzs
    • RotPwr
    • LSShftFxa
    • LSShftFys
    • LSShftFzs

• SrvD_UpdateStates

  • SrvD_CopyInput

  • SrvD_Input_ExtrapInterp

  • SrvD_UpdateDiscState

  • SrvD_Input_ExtrapInterp

  • DLL_controller_call

  • Torque_UpdateStates

  • Pitch_UpdateStates

  • Yaw_UpdateStates

  • TipBrake_UpdateStates

  • ! Blade Nacelle Tower Platform StrucCtrl，但是NumSStC = 0 这部分代码跳过

  • ! Compute ElecPwr and GenTrq for controller (and DLL needs this saved):

    • CalculateTorque

CalcOutputs_And_SolveForInputs

• SolveOption2

  • AD_CalcOutput
  • SrvD_CalcOutput
  • ED_InputSolve ED接收AD SD输入

    • BladeLoads.force/moment

    • towerloads.force/moment

    • GenTrq

    • HSSBrTrqC

    • BlPitchCom

    • YawMom

    • StrucCtrl loads 暂时没有

    • NacelleLoads.force/moment

• Transfer_SrvD_to_SD_MD

• ! 都在ifw更新后

• SrvD_InputSolve

• ResetRemapFlags

ed 到Bladed dll的输出：

   y%RotPwr    = m%AllOuts(  RotPwr)*1000.
   DO K=1,p%NumBl
      y%RootMxc(K)   = m%AllOuts( RootMxc(K) )*1000.
      y%RootMyc(K)   = m%AllOuts( RootMyc(K) )*1000.
   END DO
   y%YawBrTAxp = m%AllOuts( YawBrTAxp)
   y%YawBrTAyp = m%AllOuts( YawBrTAyp)
   !y%LSSTipPxa = m%AllOuts( LSSTipPxa)*D2R ! bjj: did this above already

   y%LSSTipMxa = m%AllOuts(LSShftMxa)*1000.
   y%LSSTipMya = m%AllOuts(LSSTipMya)*1000.                ! Rotating hub My (GL co-ords) (Nm)
   y%LSSTipMza = m%AllOuts(LSSTipMza)*1000.                ! Rotating hub Mz (GL co-ords) (Nm)
   y%LSSTipMys = m%AllOuts(LSSTipMys)*1000.                ! Fixed hub My (GL co-ords) (Nm)
   y%LSSTipMzs = m%AllOuts(LSSTipMzs)*1000.                ! Fixed hub Mz (GL co-ords) (Nm)
   y%YawBrMyn  = m%AllOuts( YawBrMyn)*1000.                ! Yaw bearing My (GL co-ords) (Nm) !tower accel
   y%YawBrMzn  = m%AllOuts( YawBrMzn)*1000.                ! Yaw bearing Mz (GL co-ords) (Nm)
   y%NcIMURAxs = m%AllOuts(NcIMURAxs)*D2R                  ! Nacelle roll    acceleration (rad/s^2) -- this is in the shaft (tilted) coordinate system, instead of the nacelle (nontilted) coordinate system
   y%NcIMURAys = m%AllOuts(NcIMURAys)*D2R                  ! Nacelle nodding acceleration (rad/s^2)
   y%NcIMURAzs = m%AllOuts(NcIMURAzs)*D2R                  ! Nacelle yaw     acceleration (rad/s^2) -- this is in the shaft (tilted) coordinate system, instead of the nacelle (nontilted) coordinate system
   y%LSShftFxa = m%AllOuts(LSShftFxa)*1000.                ! Rotating low-speed shaft force x (GL co-ords) (N)
   y%LSShftFys = m%AllOuts(LSShftFys)*1000.                ! Nonrotating low-speed shaft force y (GL co-ords) (N)
   y%LSShftFzs = m%AllOuts(LSShftFzs)*1000.                ! Nonrotating low-speed shaft force z (GL co-ords) (N)

 m%AllOuts(LSShftMxa) =  DOT_PRODUCT( MomLPRot, m%CoordSys%e1 )
 m%AllOuts(   RotPwr) = ( x%QDT(DOF_GeAz) + x%QDT(DOF_DrTr) )*m%AllOuts(LSShftMxa)
 
DO K=1,p%NumBl
	m%AllOuts( RootMxc(K) ) = DOT_PRODUCT( MomH0B(:,K), m%CoordSys%i1(K,:) )
	m%AllOuts( RootMyc(K) ) = DOT_PRODUCT( MomH0B(:,K), m%CoordSys%i2(K,:) )
END DO !K

m%AllOuts(YawBrTAxp) =  DOT_PRODUCT( LinAccEO, m%CoordSys%b1 )
m%AllOuts(YawBrTAyp) = -DOT_PRODUCT( LinAccEO, m%CoordSys%b3 )

y%LSSTipPxa = x%QT (DOF_GeAz) + x%QT  (DOF_DrTr) + p%AzimB1Up + PiBy2
m%AllOuts(LSSTipPxa) = y%LSSTipPxa*R2D


m%AllOuts(LSShftMxa) =  DOT_PRODUCT( MomLPRot, m%CoordSys%e1 )
m%AllOuts(LSSTipMya) =  DOT_PRODUCT( MomLPRot, m%CoordSys%e2 )
m%AllOuts(LSSTipMza) =  DOT_PRODUCT( MomLPRot, m%CoordSys%e3 )
m%AllOuts(LSSTipMys) = -DOT_PRODUCT( MomLPRot, m%CoordSys%c3 )
m%AllOuts(LSSTipMzs) =  DOT_PRODUCT( MomLPRot, m%CoordSys%c2 )

m%AllOuts( YawBrMyn) = -DOT_PRODUCT( MomBNcRt, m%CoordSys%d3 )
m%AllOuts( YawBrMzn) =  DOT_PRODUCT( MomBNcRt, m%CoordSys%d2 )


m%AllOuts(NcIMURAxs) =      DOT_PRODUCT(                 AngAccER  , m%CoordSys%c1 )*R2D
m%AllOuts(NcIMURAys) = -1.0*DOT_PRODUCT(                 AngAccER  , m%CoordSys%c3 )*R2D
m%AllOuts(NcIMURAzs) =      DOT_PRODUCT(                 AngAccER  , m%CoordSys%c2 )*R2D

m%AllOuts(LSShftFxa) =  DOT_PRODUCT(  FrcPRot, m%CoordSys%e1 )
m%AllOuts(LSShftFys) = -DOT_PRODUCT(  FrcPRot, m%CoordSys%c3 )
m%AllOuts(LSShftFzs) =  DOT_PRODUCT(  FrcPRot, m%CoordSys%c2 )

self.avr_swap[10] = self.demanded_pitch as f32;         self.avr_swap[34] = self.gen_state as f32; 1

49
50
51
报错信息

        self.avr_swap[60] = self.num_blades as f32;

        self.avr_swap[68] = self.root_mxc[0] as f32;

        self.avr_swap[69] = self.root_mxc[1] as f32;

        self.avr_swap[70] = self.root_mxc[2] as f32;

78 默认是4

94

初始化在swap，会被外层覆盖