Passivity Control for Hybrid Simulations of Satellite...

Passivity Control for Hybrid Simulations of Satellite Docking

Rainer Krenn, Klaus Landzettel, Toralf

Boge, Melak Zebenay

Institute of Robotics and Mechatronics

& German Space Operations Center German Aerospace Center (DLR), Oberpfaffenhofen, D-82234 Wessling

ICRA11 Space Robotics Workshop

May 13, 2011, Shanghai, China

2ICRA11 Space Robotics Workshop, Rainer Krenn, May 13, 2011

Contents of Workshop Contribution

Simulation activities in the context of On-Orbit ServicingFocus: Contact dynamics simulation

On-orbit servicing projects and contributionsCX-OLEV & Smart-OLEV

Modeling and simulation in softwareContact dynamics modelingDocking simulation results

Hybrid simulation using EPOS RvD

facilityIntroduction of hybrid simulation methodIntroduction of EPOS facilityContact dynamics simulation with EPOS

Numerical stability aspectsPassivity control


On-Orbit Servicing Projects –

OLEV Type Missions (Orbital Live Extension Vehicle)

Projects involved:CX-OLEV (ESA, Dutch Space & contractors)Smart-OLEV (Kayser-Threde, SSC, Senser

& contractors)Clients: Telecom satellites in GEOObjectives: Offer commercial services

Satellite operators (Eutelsat, Optus)Satellite live extensionFleet management

DLR contributionsRobotic docking tool (Capture Tool)Rendezvous and docking control algorithms (image processing) Docking simulations (S/W)


OLEV Servicing

Satellite

& Docking Payload

(Smart)

1

3

2

554

4

4

1) Capture Tool (CT)2) CT Deployment Mechanism (CDM)3) Target Illumination System (TIS) (2x)

4) Camera System (Far-, Mid-, Near-Range

Cameras with Electronics) (2x)5) Docking Payload Control Unit (2x)


Capture Tool for Non-Cooperative Targets


Docking Scenario

for Multi-Body System Simulation

ClientHot Bird series, with flexible solar wingsOrbit: GEOAOCS deactivated, momentum wheels loadedAerojet/Marquardt R4D nozzle, slightly tilted1194 mm launch adapter

Smart-OLEVOrbit: GEO + 2.178 m (R-bar)AOCS active75 mm

lateral misalignment,

2°

attitude error

in all axesFlexible Capture Tool deployment mechanism

|v| ≤

4 mm/s deployment/retraction velocityCapture Tool

Radial laser distance sensorContact switchesOperational Locking Crown

Contact sensitive bodies:Nozzle vs. Capture Tool + Locking CrownLaunch adapter vs. Client Support Brackets

Smart-OLEV

Support Brackets

Flexible BoomCapture Tool

Nozzle

Launch Adapter

Client Spacecraft


Contact

Dynamics Model (3D, Multi-Point)

Surface meshingBounding volume tree generation Contact area detection algorithm

Elastic foundation model for contact force computation Friction model with stick-slip


0 50 100 150 200 250 300-200

-100

0

100

200

300

400

Simulation Time [s]

App

lied

Forc

e [N

]Lo

ckin

g C

row

nLo

ckin

g C

row

n R

F

W02

xyz

0 50 100 150 200 250 300-5

0

5

10

15

Simulation Time [s]

App

lied

Torq

ue [N

m]

Lock

ing

Cro

wn

Lock

ing

Cro

wn

RF

W02

xyz

0 50 100 150 200 250 300-0.06

-0.05

-0.04

-0.03

-0.02

-0.01

0

0.01

0.02

0.03

Simulation Time [s]

Late

ral P

ositi

on E

rror [

m]

Clie

nt C

OM

- S

mar

t CO

MS

mar

t GR

F

W02

yz

0 50 100 150 200 250 300-0.01

0

0.01

Simulation Time [s]

Ela

stic

Def

orm

atio

n [-]

CD

M S

egm

ents

Ben

ding

W02

0 50 100 150 200 250 300-5

0

5x 10

-3

Simulation Time [s]

Ela

stic

Def

orm

atio

n [-]

CD

M S

egm

ents

Tors

ion

W02

0 50 100 150 200 250 300-5

0

5

10x 10

-3

Simulation Time [s]

Ela

stic

Def

orm

atio

n [m

]C

DM

Seg

men

tsA

xial

W02

0 50 100 150 200 250 300-100

-50

0

50

Simulation Time [s]

App

lied

Forc

e - C

SB

1 [N

] W02

0 50 100 150 200 250 300-100

-50

0

50

Simulation Time [s]

App

lied

Forc

e - C

SB

2 [N

] W02

0 50 100 150 200 250 300-100

-50

0

50

Simulation Time [s]

App

lied

Forc

e - C

SB

3 [N

] W02


Hybrid Simulation Method

1.

Numerical simulation of satellite trajectory (S/W)Free floating dynamics of satellites, 6 DOFMicro-gravity / 0-gravity conditionsLarge flexible structures, solar panelsMomentum wheels, reaction wheels, thrusters

2.

Physical-mechanical display of computed trajectory

using industrial robotsAttachment of satellite mockupsPerform physical contact (capture tool, nozzle) Use of missions specific mechanisms,

actuators and sensors (capture tool, GNC-sensors)

3.

Sensor data feedback from facility specific sensorsRobot joint sensors actual trajectoryForce torque sensors contact dynamics purposes


Hybrid Simulation Facility EPOS (European Proximity Operations Simulator)

1.

Development system for dynamics models: Matlab/SimulinkReal-time target for numerical simulation: VxWorks

2.

Robotic system for trajectory emulation: (Kuka, RSI, position controlled)3.

Force/torque sensor feedback: Schunk/DLR FTS (“Compliance”)

EPOS has a long history as open-loop satellite rendezvous simulator (ATV sensor verification).But: Is EPOS applicable for closed-loop contact dynamics simulations?


Hybrid Simulation of Docked Configuration

Sat 1 Sat 2

FTS

Sat 1 Sat 2

System to be simulated“Ideal System”

Hybrid simulation setup

k

k

Numerical Simulation

Robotic System

Free floating rigid bodiesSpring at natural lengthv = 0


Experimental Setup “Docked Configuration”

Simulating the elastically fixed configuration

(1 DOF) using single robot and

one-end supported aluminum sheet metal1.

Open hybrid simulation loop2.

Initial robot/satellite position with bent beam3.

Sensor calibration/reset (F = 0)4.

Close hybrid simulation loop again

Easy mechanical setupNearly linear stiffness and low dampingEasy stiffness adaptation by changing contact point distance from support point

(1500 N/m …

3000 N/m)Low risk of mechanical damage during experiment


Observations

Expected: Stationary system at restObserved: Unstable system

Oscillating systemIncreasing amplitudes

0 50 100 150 200-30

-20

-10

0

10

20

30

Simulation Time [s]

App

lied

Con

tact

For

ce [N

]

0 50 100 150 200-0.04

-0.02

0

0.02

0.04

Simulation Time [s]

Pos

ition

[m]

RobotSimulation

0 50 100 150 200

-0.03

-0.02

-0.01

0

0.01

0.02

0.03

Simulation Time [s]

Vel

ocity

[m/s

]

RobotSimulation


Sources of Excitation? ---

Active Elements?

Simulation sampling rate not adequate for system dynamics?250 Hz >> eigen-frequency of contact dynamics problem

Force-Torque sensor errors and noise? Calibration error may cause change of static equilibrium positionMean of noise is zeroBut: May initiate system instability if active elements are in the loop

Passivity of robot system to be investigated

Numerical Simulation

Robotic System

Physical Contact

FT-Sensor

Simv Robv

FTSF FTSFOne-Port

NetworkOne-Port

NetworkTwo-Port

Network


Observed Energy ErrorRobotic

System

Ctrlv Robv

FTSF FTSF

0

0

k

Error out i in ii

k

FTS i Rob i Ctrl ii

E dt P t P t

dt F t v t v t

0 50 100 150 200-20

0

20

40

60

80

100

120

Simulation Time [s]

Ene

rgy

Erro

r in

Sys

tem

[J]

0 50 100 150 200-0.06

-0.04

-0.02

0

0.02

0.04

0.06

0.08

0.1

0.12

Simulation Time [s]

Pow

er E

rror o

f Rob

ot [W

]


Transfer Function of Robot System

123 123.05 123.1 123.15 123.2 123.25 123.3 123.35 123.40.0153

0.0153

0.0154

0.0154

0.0154

0.0154

0.0154

0.0154

0.0154

0.0154

0.0154

Simulation Time [s]

Pos

ition

[m]

RobotSimulation

70 ms

70 ms

76.6 76.62 76.64 76.66 76.68 76.7-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1x 10-5

Simulation Time [s]

Pos

ition

[m]

RobotSimulation

24 ms

Electro-mechanical systemAmplitude response of robot armPhase response of robot arm

Robot control system with preprocessing of external commands

LatencySmoothing of input signals


Passivity ControlRobotic

System

Ctrlv Robv

FTSF FTSF

PC

SimF

Simv

11 12

11

1

11

1

; ; 0;

0; 0; 0

;;

Impedance causality:

Admittance causality:

Error iv Error i Error i

Sim iSim i FTS i i Sim i i

Error i

Error iF Error i

Ctrl i Sim i i FTS i i

E tk E t E t

dt v tF t F t t v t tE t

E tk E t

dtv t v t t F t t

121

1

; 0

0; 0; 0

Error iFTS i

Error i

E tF t

E t

PID type Energy ControlHybrid simulation of space robot applications

[Krenn, Schäfer, AIAA GNC 1999]

Time Domain Passivity ControlHaptic

devices[Hannaford, Ryu, ICRA 2001]

Dissipate energy errorMaintain system dynamics


0 100 200 300 400 500 600-2

-1

0

1

2x 10-3

Simulation Time [s]

Vel

ocity

[m/s

]

RobotSimulation

0 100 200 300 400 500 600-1.5

-1

-0.5

0

0.5

1

1.5

Simulation Time [s]

App

lied

Con

tact

For

ce [N

]

Experiment Results Using PID Type Energy Control

0 100 200 300 400 500 600-0.08

-0.06

-0.04

-0.02

0

0.02

0.04

0.06

0.08

Simulation Time [s]

Con

trolle

r For

ce [F

]

0 100 200 300 400 500 600-5

0

5

10

15x 10-3

Simulation Time [s]

Ene

rgy

Erro

r in

Sys

tem

[J]


0 50 100 150 200 250-0.2

-0.1

0

0.1

0.2

Simulation Time [s]

Con

trolle

r For

ce [N

]

Experiment Results Using Time Domain Passivity Control

0 50 100 150 200 250-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

Simulation Time [s]

App

lied

Con

tact

For

ce [N

]

0 50 100 150 200 250-5

0

5

10

15x 10-5

Simulation Time [s]

Ene

rgy

Erro

r in

Sys

tem

[J]

0 50 100 150 200 250-1.5

-1

-0.5

0

0.5

1

1.5x 10-3

Simulation Time [s]

Vel

ocity

[m/s

]

RobotSimulation


Conclusion & Future Work

Meaningful results for design phases A/B using software simulation.Hybrid simulation required for phases C/D/E

Real-time simulation requirementHardware-in-the-loop simulation requirements

Robotic systems are not passive Numerical instabilities of closed-loop simulationPassivity control solutions successfully implemented and tested at EPOS

PID energy control approachTime domain passivity control

Future work:Passivity tests at higher contact stiffnessTest of system transparency (accuracy of contact dynamics simulation)

Passivity Control for Hybrid Simulations of Satellite...

Documents

Transcript of Passivity Control for Hybrid Simulations of Satellite...