Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Real Time Simulation of Flexible Aerospace Vehicles Using Simulink and RT-Lab
Yadunath Gupta, IIT Kharagpur
Pulak Halder, Research Centre Imarat
Siddhartha Mukhopadhyay, IIT Kharagpur
23-04-2015 1© 2015. All Rights Reserved
Agenda
• Flexible aerospace vehicles
• Controller design
• Challenges
• Strategy
• Simulation and results
• Conclusion
23-04-2015 2© 2015. All Rights Reserved
Flexible Aerospace Vehicles
• Definition: Vehicles that undergo structural deformation during flight
• High length to diameter ratio
• Aerodynamically more efficient
• Lower radar visibility
23-04-2015 3© 2015. All Rights Reserved
Effects of Flexibility
• Corrupts the feedback signals from sensors
• Vibration frequencies can interfere with control frequencies.
• May lead to self excited divergent oscillations
Proper controller design and simulation of vehicle flight dynamics is necessary to ensure stability
23-04-2015 4© 2015. All Rights Reserved
Controller Design
• Gain stabilization for high frequency modes
• Phase stabilization for lower frequency modes
23-04-2015 5© 2015. All Rights Reserved
Non-Real Time vs Real Time Simulation
• Non-real time:
– Generally used for validation of algorithms
• Real time:
– Essential for validation of controllers which are designed using phase stabilization.
– Can simulate effects of computational delays
– Enables testing of hardware components
23-04-2015 6© 2015. All Rights Reserved
Challenges
• Complexity and order of the model is large
• High computational load
• Model compatibility with real time simulator
• Support for IO cards / communication interfaces in hardware-in-loop simulation
• Monitoring and debugging
23-04-2015 7© 2015. All Rights Reserved
Strategy
• MATLAB and Simulink
• RT LAB
• Distributed architecture
• Multi-rate simulation
23-04-2015 8© 2015. All Rights Reserved
Tactical Aerospace Vehicle and its Model in MATLAB®/Simulink
23-04-2015 9
Atmospheric ModelVehicle Dynamics
Model (6DoF)Guidance & ControlNavigation
© 2015. All Rights Reserved
Control SurfaceWing
Navigation Guidance & Control
Motor (propulsion)
Navigation Guidance ControlAirframe &
Propulsion
Vehicle
Dynamics
~
Vx~
X~
Acceleration, Rates
ActuatorPlant
Advantages:
Conversion from Non Real-time to Real-Time is very easy
Easy to Implement Communication interfaces (MIL std 1553B, SharedMemNet, ADC/DAC etc.)
Easy to trace the signal, Introduction of disturbance
Validation of model developed by external Agencies (Academic institution/Collaborating Org)
has become easier (Plug & Play).
Vehicle Dynamics Model
23-04-2015 10
PY_Bending
3
Mslp_INS _RGP_PY
2
axyzpqrdot _RFlex
1
In1
In2Out1
-1
1/57 .3
Sarea
1/57 .3
Rigid _6DOF MODEL
delta_p
delta_y
delta_r
pa
alpha
beta
dp
mac
Phi_actual
time
qr_RFlex
ThMscgIyyIxx
axyz _pqrdot_R
FLEX _6DOF_Model
Thrust
mass
xcg
Iyy
axyz _pqrdt_Rigid
tm
Alp_beta
AlphaT
mac
DynPr _sArea
Delta_PY
Vel_b
axyzpqrdot _RFlex
Mslp_INS_RGP_PY
qr_RFlex
PY _Bending
Vel_b
6
time
5
Atmos_para
4
delta _r
3
delta _p
2
delta _y
1
Rigid Body Model
Flexibility Model
© 2015. All Rights Reserved
Flexibility Model
23-04-2015 11
PY_Bending
4
qr_RFlex
3
Mslp_INS _RGP_PY
2
axyzpqrdot _RFlex
1
Vel_b1
12
Eng_Acc _Gyro _modfLoc
omega_flex _col
gen_mass_col
gm_dot_col
tm _col
xcor
xcna
phy 1
phydsh 1
phy 2
phydsh 2
phy 3
phydsh 3
tm
Mshp_Mslp_TE
Mshp_INS
Mslp_INS
Mslp_RG
wz_GM_GM_dt
rphy
rphydsh
ModeShpSlp _calc
Mslp _INS
Mslp_RG
FCoe _qfdt
FCoe _ayf
FCoe _qyddt
xcg_E_T
wz_mz_mz_dt
Mshp _INS
Mshp _Mslp_TE
Flex _Yaw_Plane
ay _R
r_dot_rigid
Thrust
Vx
Vyb
mass
Izz
beta
deltaY
r_RFlex
fy _ins_rdot_cg
Mslp_INS_RGP_Y
YawBending
Flex_Pitch _Plane
az_R
q_dot_rigid
Thrust
Vx
Vzb
mass
Iyy
Alpha
deltaP
fz_ins_qdot_cg
Mslp_INS _RGP_P
q_RFlex
PitchBending
rphy
rphydsh
mac
QS
alphaT
Vel_b
xcna
mach_col
Cna1
Cna2
Cna3
xcgE
FCoe_qyddt
FCoe_ayf
FCoe_qfdt
force _coeff
xgimb
Cna2
Cna1
gm _dot _col
gen _mass_col
omega _flex _col
mach _col
Cna3
Eng _Acc_Gyro_modfLoc
xcna
phydsh 3
phy 3
phydsh 2
phydsh 1
phy 2
phy 1
xcor
tm_col
sitvl
xcna
Vel_b
12
Delta _PY
11
DynPr_sArea
10mac
9
AlphaT
8
Alp _beta
7
tm
6
axyz_pqrdt _Rigid
5
Iyy
4
xcg
3
mass
2
Thrust
1
Mode shape & slope ComputationFlexibility Rate & Lateral Acc. Computation in Pitch/Yaw plane
Force Coef. Computation
© 2015. All Rights Reserved
Pitch Plane Flexibility Model
23-04-2015 12
PitchBending
4
q_RFlex
3
Mslp _INS_RGP _P
2
fz_ins_qdot _cg
1
Vx
Alpha
Thrust
deltaP
qy
qy _dot
qy _ddot
qy
qy_dot
qy_ddot
Thrust
mt
Vy _dot_f -u
-u
-u
Thrust 1
3
TPZ 12
In Out
qy
qy_dot
qy_ddot
Thrust
delta_P
Iyy
Vy
r_dot_f
r_rdot
-K-
-K-
Flx_flag _P
Flx_flag _P
-K-
Flx_flag _P
Flx_flag _P
Flx_flag _P
Mslp _RG
Mslp_INS
Mshp_INS
deltaP
9
Alpha
8
Iyy
7
mass
6
Vzb
5
Vx
4
Thrust
3
q_dot _rigid
2
az_R
1
Generalized Coordinate
ComputationFlexibility Lateral Acc. Computation
Flexibility Rate Computation
© 2015. All Rights Reserved
Simulation Scenario
• Scenario of a flight test which failed due to excessive bending oscillations was reproduced
• Flight response of the launch vehicle with rigid body model and flexible body model was analysed
• Updated controller was validated
23-04-2015 13© 2015. All Rights Reserved
Vehicle Response with Rigid-Body Model
23-04-2015 14© 2015. All Rights Reserved
Vehicle Response with Flexible-Body Model
23-04-2015 15© 2015. All Rights Reserved
Vehicle Response with Flexible-Body Model and Updated Controller
23-04-2015 16© 2015. All Rights Reserved
Conclusion
• MATLAB and Simulink tools enabled to develop a complete model of the dynamics of a flexible aerospace vehicle
• RT LAB helped in real time simulation of the vehicle using Simulink models, with hardware in loop
• Block based modelling technique eased the process of debugging, monitoring and modifying the model, and allowed for flexible degrees of freedom to be added to the existing rigid body model
23-04-2015 17© 2015. All Rights Reserved
Thank you!
23-04-2015 18© 2015. All Rights Reserved