SDL Control of the UltraLITE Precision Deployable Test Article - - PowerPoint PPT Presentation

sdl
SMART_READER_LITE
LIVE PREVIEW

SDL Control of the UltraLITE Precision Deployable Test Article - - PowerPoint PPT Presentation

SDL Control of the UltraLITE Precision Deployable Test Article Using Adaptive Spatio-Temporal Filtering Based Control Albert B. Bosse Keith K. Denoyer R. Scott Irwin Thomas D. Sharp Air Force Research Laboratory Stuart J. Shelley Kirtland


slide-1
SLIDE 1

3/3/99, Page SDL-1

SDL

Control of the UltraLITE Precision Deployable Test Article Using Adaptive Spatio-Temporal Filtering Based Control Albert B. Bosse Thomas D. Sharp Stuart J. Shelley

Sheet Dynamics, Ltd. Cincinnati, OH

Keith K. Denoyer

  • R. Scott Irwin

Air Force Research Laboratory Kirtland Air Force Base

slide-2
SLIDE 2

3/3/99, Page SDL-2

SDL

Overview

  • UltraLITE Deployable Optical Telescope program
  • DOT test beds

– Mirror Mass Simulator – PDOS – DOT BGD

  • Active structural control issues
  • Spatio-Temporal Filtering (STF)
  • STF based structural control
  • PDOS test experience
slide-3
SLIDE 3

3/3/99, Page SDL-3

SDL

Acknowledgements

  • Phase II SBIR award from Ballistic Missile

Defense Organization (BMDO)

  • Contract managed by and technical collaboration

with Air Force Research Lab - Kirtland AFB

slide-4
SLIDE 4

3/3/99, Page SDL-4

SDL

Air Force Research Laboratory Deployable Space Telescope Concept

  • Large

aperature/resolution through deployable, sparse, optical array

  • Deployable primary

mirrors

  • Golay 6 configuration
  • Telescoping seconday

tower

slide-5
SLIDE 5

3/3/99, Page SDL-5

SDL Deployable Optics Concepts Evaluation

Experimental Test Beds

  • Mirror Mass Simulator (MMS)
  • Precision Deployable Optics Structure (PDOS)
  • Deployable Optical Telescope Brassboard Ground

Demonstration (BGD)

slide-6
SLIDE 6

3/3/99, Page SDL-6

SDL

Preliminary DOT Evaluation - Mirror Mass Simulator

  • Mirror Mass

Simulator mounted to

  • ptics bench
  • 3 interferometer

displacement sensors

  • 3 piezo stack

actuators

  • Electromagnetic

disturbance shaker

slide-7
SLIDE 7

3/3/99, Page SDL-7

SDL

Precision Deployable Optical Structure

Interface Deployable Boom Actuation Subsystem Optical Subsystem & Truth Sensor Mirror Simulator Granite Slab

slide-8
SLIDE 8

3/3/99, Page SDL-8

SDL

Integrated Technology in Simulated Space/Ops Environment Supporting SBL, Global Virtual Presence, and Tactical Imaging Missions

Composite/Glass Hybrid Mirrors 3m Deployable Secondary Tower Secondary Mirror Deployable Reaction Plate Precision Mechanisms Aft Optics Adaptive Optics Wavefront Sensors Optical Bench

1.5m Deployable Telescope

Deployable Optical Telescope

Brassboard Ground Demonstration

Optical Test Bench:

Finite Conjugate Scene Center of Curvature Sensor Laser Metrology

Test Tower

Deployable Telescope

slide-9
SLIDE 9

3/3/99, Page SDL-9

SDL

  • Precision Deployable Optical Structure (PDOS):

Achieve 30 nanometers or less RMS value for relative displacement between the granite slab and the mirror mass simulator

  • Deployable Optical Telescope (DOT)

(1) Maintain the position of the primary mirror segments within: Piston: ± 14 nanometers error per segment Tilt: ± 95 nanoradians error per segment (2) Maintain the position of the secondary mirror within:

Decenter ±50 microns

Piston: ± 4 microns Tilt: ± 20 microradians

Requirements

SDL’s primary mission is to provide a Vibration Control System that will assist the Optical Control System in meeting the DOT mirror positioning requirements

slide-10
SLIDE 10

3/3/99, Page SDL-10

SDL

Deployable Optics - Jitter Requirements

  • Disturbances

– torque wheel actuators – slewing – space based laser

  • Vibration Control

– isolation – passive vibration control – high bandwidth position control – active vibration control

slide-11
SLIDE 11

3/3/99, Page SDL-11

SDL

Active Structural Vibration Control Issues

  • Modeling - accurate and complete dynamic

models of complex “real-world” systems are difficult to obtain.

  • Time Variance - Often, by the time you’ve got the

model the system has changed - It’s a moving target.

– System dynamics - temperature, load, wear, damage – Discrete failures - sensors, actuators, signal conditioning

  • Computational burden
slide-12
SLIDE 12

3/3/99, Page SDL-12

SDL

Spatio-Temporal Filter Based Control

10 20 30 40 50 60 70 80 90 100 10
  • 2
10
  • 1
10 10 1 10 2 10 20 30 40 50 60 70 80 90 100 10 -2 10 -1 10 0 10 1 10 2 10 20 30 40 50 60 70 80 90 100 10
  • 2
10
  • 1
10 10 1 10 2 10 20 30 40 50 60 70 80 90 100 10
  • 2
10
  • 1
10 10 1 10 2 10 20 30 40 50 60 70 80 90 100 10
  • 2
10
  • 1
10 10 1 10 2 10 20 30 40 50 60 70 80 90 100 10
  • 2
10
  • 1
10 10 1 10 2

Uncontrolled Response Controlled Response Uncontrolled Modal Responses Extracted with STF Controlled Modal Responses

slide-13
SLIDE 13

3/3/99, Page SDL-13

SDL

Modal Coordinate Transformation Uncouples System into SDOF Modes

Mx Cx Kx f

  • +

+ =

  • m

c k f

T

L N M M M O Q P P P

+L

N M M M O Q P P P

+L

N M M M O Q P P P

=

  • η

η η Φ

x t t t

r r r N

b g b g b g

= =

=

∑φ η

η

1

Φ

Modal Coordinate Transformation

slide-14
SLIDE 14

3/3/99, Page SDL-14

SDL

STF Origin - Modal or Spatial Filtering

ψ φ

i T r

i r i r = ≠ = = 1

ψ ψ φ η ψ φ η η

i T i T r r r N i T i i i

x t t t t ( ) ( ) ( ) ( ) = = =

=

1

Spatial filter vector ψ Extract single mode response from measured response

slide-15
SLIDE 15

3/3/99, Page SDL-15

SDL

Spatio-Temporal Filtering

  • η

ψ

k T k

x =

  • η

ψ

k T k k k Nt

x x x =

R S | | T | | U V | | W | |

− − 1

  • Spatial filter

estimate of at time k

η

Spatio-Temporal filter estimate of at time k

η

slide-16
SLIDE 16

3/3/99, Page SDL-16

SDL

Spatio-Temporal vs. Modal Filtering

  • FIR or “all-zero” filter on each channel
  • Pole-zero cancellation & preferential pass filter

– fewer sensors required

  • Inherent estimation of modal velocity
  • Compensation for filter delays, sensor & signal

conditioning dynamics

  • Non-homogeneous sensor suites - piezo patches,

accelerometers, etc.

slide-17
SLIDE 17

3/3/99, Page SDL-17

SDL Adaptive Calculation of STF Coefficients

using Reference Model Approach

  • Know only poles of controlled modes
  • Don’t know

– mode shapes – modal scaling factors (modal mass) – modal participation vectors – anything about uncontrolled modes (not even poles)

slide-18
SLIDE 18

3/3/99, Page SDL-18

SDL Adaptive Calculation of STF Coefficients

using Reference Model Approach

η η

λ k r k r T k

z l f

+ =

+

1

b g b g

η η η η

λ λ k r k r k k rN k rN k N

z f z f

i i i

+ +

= + = +

1 1 1 1 1

b g b g b g b g b g b g

  • η

η η η

k r T k r k rN T k r

l l

i

b g b g b g

=

R S | T | U V | W |

=

1

  • +

SDOF (Single Mode) Reference Model

slide-19
SLIDE 19

3/3/99, Page SDL-19

SDL Adaptive Calculation of STF Coefficients

using Reference Model Approach

e l x x x l x x

k k r k T k r T k k k Nto T k k Nto k r

= − = −

R S | | T | | U V | | W | |

=R

S T U V W

− −

R S | | T | | U V | | W | |

− − −

η η η ψ ψ η

b g

  • 1
slide-20
SLIDE 20

3/3/99, Page SDL-20

SDL

STF Based Modal Velocity Feedback Control

f v

c i i i i

b g b g b g b g

=

  • η α

f l

c i i i i

b g b g b g b g

=

  • η α

Modal Coordinate Velocity Estimate Scalar Feedback Gain Force Vector Control Command Vector for i’th mode Estimated Modal Participation Vector is Ideal Force Vector

slide-21
SLIDE 21

3/3/99, Page SDL-21

SDL

Initial Mirror Mass Simulator Control Experiments

  • STF based velocity feedback
  • 3 inputs, 3 outputs, 5 controlled modes
  • Random disturbance excitation
  • 1 1/2 days to implement

– familiarization with test bed – all system ID – control implementation and testing

slide-22
SLIDE 22

3/3/99, Page SDL-22

SDL

Implementation of STF Based Control

Actuators Amplifiers Sensors Signal Cond Structure ADC's DAC's

ψ

  • Ref. Model

Random Excitation

+

e

Controller

Control Command Control/Excitation Command

  • η

r

b g

  • η

dSpace System

l l

  • η
slide-23
SLIDE 23

3/3/99, Page SDL-23

SDL

Preliminary DOT Evaluation - Mirror Mass Simulator

  • Mirror Mass

Simulator mounted to

  • ptics bench
  • 3 interferometer

displacement sensors

  • 3 piezo stack

actuators

  • Electromagnetic

disturbance shaker

slide-24
SLIDE 24

3/3/99, Page SDL-24

SDL

Initial Mirror Mass Simulator Control Experiments

50 100 150 200 250 300 350 400 10

  • 2

10

  • 1

10 10

1

Frequency - Hertz Interferometer 1 versus Disturbance Force Amplitude - Displacement/Force

slide-25
SLIDE 25

3/3/99, Page SDL-25

SDL

Initial Mirror Mass Simulator Control Experiments

50 100 150 200 250 300 350 400 10
  • 2
10
  • 1
10 10 1

Frequency - Hertz Interferometer 2 versus Disturbance Force Amplitude – Displacement/Force

slide-26
SLIDE 26

3/3/99, Page SDL-26

SDL

Initial Mirror Mass Simulator Control Experiments

50 100 150 200 250 300 350 400 10

  • 2

10

  • 1

10 10

1

Frequency - Hertz Amplitude - Displacement/Force Interferometer 3 versus Disturbance Force

slide-27
SLIDE 27

3/3/99, Page SDL-27

SDL

Precision Deployable Optical Structure

Proved Deployment, Acquisition, Maintenance and Control System for a 2m Optical Segment

Mirror Inertial Simulator Gravity Off-Load Composite Boom Reference Bench Back-Up Structure

slide-28
SLIDE 28

3/3/99, Page SDL-28

SDL

Precision Deployable Optical Structure

slide-29
SLIDE 29

3/3/99, Page SDL-29

SDL

Precision Deployable Optical Structure

slide-30
SLIDE 30

3/3/99, Page SDL-30

SDL

PDOS Frequency Response Function

10 20 30 40 50 60 70 80 90 100

  • 20
  • 10

10 20 30 40

TEST MODEL

Interferometer 1 / PZT 2 FRF

  • 5 Hertz boom mode only

mode in low frequency range

  • 1.6 Hertz slab mode not

apparent - must treat as a disturbance

slide-31
SLIDE 31

3/3/99, Page SDL-31

SDL

STF Control PDOS 5 Hz Boom Mode

2 4 6 8 10 12 14 16 18 20

  • 1
  • 0.8
  • 0.6
  • 0.4
  • 0.2

0.2 0.4 0.6 0.8 1

OPEN LOOP

2 4 6 8 10 12 14 16 18 20

  • 1
  • 0.8
  • 0.6
  • 0.4
  • 0.2

0.2 0.4 0.6 0.8 1

CLOSED LOOP

slide-32
SLIDE 32

3/3/99, Page SDL-32

SDL

Open and Closed Loop Interferometer PSD’s

2 4 6 8 10 12 14 16 18 50 100 150 2 4 6 8 10 12 14 16 18 50 100 150

MAG (dB)

2 4 6 8 10 12 14 16 18 20 50 100 150

FREQUENCY (Hz)

OPEN LOOP CLOSED LOOP 20 20

slide-33
SLIDE 33

3/3/99, Page SDL-33

SDL

PDOS Ambient Vibration

10 20 30 40 50 60 70 80 90 100 10 10

1

10

2

10

3

10

4

10

5

10

6

10

7

ch 1 ch 2 ch 3

  • RMS Vibration, 0-250 Hz
  • Int 1: 171-430 nm
  • Int 2: 117-264 nm
  • Int 2: 93-239 nm
  • Resonant and forced

vibration

Resonant Frequencies

slide-34
SLIDE 34

3/3/99, Page SDL-34

SDL

Interferometer RMS Value Versus Frequency Band

5 10 15 20 25 30 35 40 45 50 10 10

1

10

2

10

3

FREQ (Hz) RMS DISPLACEMENT (nm) ch 1 ch 2 ch 3

slide-35
SLIDE 35

3/3/99, Page SDL-35

SDL

PDOS Low Frequency Control

10 20 30 40 50 60 70 80 90 100 10

  • 1

10 10

1

10

2

10

3

10

4

10

5

ch 1 ch 2 ch 3

  • RMS Vibration, 0-250 Hz
  • Assuming 50x reduction

0-13 Hz.

  • Int 1: 29.99 nm
  • Int 2: 28.51 nm
  • Int 3: 27.42 nm
  • Note the control objective is 30 nm RMS vibration levels
slide-36
SLIDE 36

3/3/99, Page SDL-36

SDL

Conclusions

  • Resonant control alone is not sufficient to meet

PDOS/DOT optical jitter control requirements

  • “High” bandwidth position control in conjunction

with resonant mode control required

  • STF based modal control is practical approach for

resonant mode control;

– Implement effective MIMO control on complex, “real- world” structures with little knowledge of dynamics – Adapts to sensor/actuator failure – Accommodates filter/signal conditioning dynamics – Easily updated to accommodate changing system dynamics (only update poles of controlled modes)