[PPT] - Air Bearing System for Space Dynamics and Control Simulations Dr. PowerPoint Presentation

SLIDE 1

Air Bearing System for Space Dynamics and Control Simulations

Dr. John T. Wen, Professor, Electrical, Computer, & Systems Eng.

David Carabis, Ph.D. Student, Mechanical, Aerospace, & Nuclear Eng. Rensselaer Polytechnic Institute September 24, 2017

Gravity Offload Testbeds for Space Robotic Mission Simulation Full-day Workshop at 2017 IROS

1

SLIDE 2

Overview

2

Develop technology and

perform ground-based testing to support operations for potential satellite servicing missions.

Problem Statement:

– Dual arm capture, transport and docking with vision/force feedback – Delay compensation in compliance control – Evaluation on air bearing testbed – Evaluation on space robot dynamic simulation

SLIDE 3

Outline

3

Air bearing testbed
Outer loop vision and force feedback
Delay compensation in compliance control
Dual arm tracking, grasping, transport, berthing
Grasp slip prevention control
Dynamic simulation
Future Work:

– Flexible material modeling, simulation, and manipulation

SLIDE 4

Air Bearing Testbed

4

Full 3-DOF motion in reduced friction environment (4’x8’)
Self contained, with on-board 3000psi air tank (reduced to 40psi)
Three flat air bearings for max load 150lb (sled mass~25lb)
Run time ~35 min, fill time ~ 50min.

SLIDE 5

Industrial Robot Controller with External Commands

5

Inner loop: robot joint controller (high rate, high gain)
Outer loop: joint increment (med rate, long latency)
Sensor loop: force/torque, relative position (low rate,

not guaranteed real-time) u = ∆q Robot

q

fq e−tds s

qc

Inner loop Outer loop

Sensor Measurements

Sensor loop

SLIDE 6

Sensor (Vision and Force) Outer Loop Control

6

u = ∆q

Robot q

fq e−tds s

qc

Fwd Kin Contact F

x

Vision F/T Sensor Force Control

xdes

Jacobian or Inv Kin Based Control Human input Visual SLAM

Fdes

SLIDE 7

Control Design Issues

7

Force/impedance control
Effect of 3me delay
Grasp stability
Redundancy resolu.on
Singularity avoidance
Collision avoidance (self collision, environment)
Joint/load flexibility
Environmental impedance (for inser.on/docking)
Sensor loop non-real-.me
Human interface

SLIDE 8

Multi-Arm Grasp

8

xc

~ pc

~ p1

~ p2

~ p1C ~ p2C

AT fs = AT f ∗

s = 0

Load Dynamics Mcαc + bc = X

i

AT

i Fi

Contact Kinematics HT

i Fi = 0

Fi = H⊥T

i

ηi

Hi = [I, 0]

Point Contact with Friction

Grasp Model Mcαc + bc = X

i

AT

i H⊥T i

ηi = Gη

G = Grasp Map

Stable Grasp Condition:

N(G) = {0} η ∈ Friction Cone

Ai =  I −(ROCpiC)× I

SLIDE 9

Outer Loop Based Motion and Force Control

9

xc

~ pc

~ p1

~ p2

~ p1C ~ p2C

Arm Kinematics Motion Control

(P or PI)

V ∗

c = −Kp(xc − x∗ c)

Force Control

(Generalized damper)

Motion/Force Control with Move-Squeeze Decomposition

V ∗

i = V ∗ im + V ∗ is

Collision avoidance: penalty function Berthing: use xc

* as virtual input for berthing control

V ∗

is = −Ks(fs − f ∗ s )

V ∗

im = AiV ∗ c ,

(HT )+AV ∗

c = 0

Gfs = Gf ∗

s = 0

Vi = Ji ˙ qi = ui

Use Vi* as virtual input for xc control

SLIDE 10

Delay Compensation in Force Control

10

Time Delays Inner-Loop Dynamics Jacobian Jacobian Pseudo-Inverse

Surface

SLIDE 11

Dual Arm Force Control with Delay Compensation

11

Impedance Controller Smith Predictor x

Gain Margin Phase Margin x

D. Kruse, J.T. Wen, “Application of the Smith-Astrom Predictor to

Robot Force Control,” IEEEE Conference on Automation Science and Engineering (CASE), Gothenburg, Sweden, Aug, 2015.

SLIDE 12

Dual Arm Load Capture (Motoman): Vision+Force

12

Motoman HSC + Robot Raconteur +

MATLAB/Simulink

ALVAR-Tag Target Tracking
Consensus Grasp + Load Transport
2ms joint increment command

and measurement

~16ms delay + 1st order filter
Vision loop ~ 30Hz

SLIDE 13

Dual Arm Load Capture (Baxter): Vision+Force

13

Same controller as Motoman with re-

tuned gains

Baxter cameras + force measurements
ROS + RR/ROS Bridge + MATLAB
~6ms joint increment command

and measurement (non-real-time)

Camera update < 2Hz (bottleneck

unclear at this point)

SLIDE 14

Desired Squeeze Force Scheduling

14

Measured Force (N)

100
50

fd (N)

120
100
80
60
40

d (m) 0.1 0.2 0.3 fd (N)

600
500
400
300
200
100

Fast motion away from contact Slow approach near contact Gentle contact first and then ramp up force (to avoid bouncing)

SLIDE 15

Consensus Dual Arm Grasping

15

V ∗

i = −Ki(xi − x∗ i ) − Kij((xi − x∗ i ) − (xj − x∗ j))

Time (s) 24.5 25 25.5 26 Force (N)

25
20
15
10
5

5 Left Arm Right Arm

(a) Centered, vision consensus.

Time (s) 24.5 25 25.5 26 Force (N)

25
20
15
10
5

5 Left Arm Right Arm

(b) Off-centered, vision consensus.

Time (s) 24.5 25 25.5 26 Force (N)

25
20
15
10
5

5 Left Arm Right Arm

(c) Centered, no vision consensus.

Time (s) 24.5 25 25.5 26 Force (N)

25
20
15
10
5

5 Left Arm Right Arm

(d) Off-centered, no vision consensus.

(+ integral terms)

SLIDE 16

Time (s) 20 40 Force (N)

150
100
50

Left Arm Right Arm Fd

Load Transport

16

Motion-induced contact force

How to choose squeeze force setpoint fs

* in the

presence of motion?

How to suppress vibration due to structural

flexibility

Contact pad vibration

Oscillatory load motion at 1 rad/sec

SLIDE 17

Slip Prevention Control

17

Objective: Maintain stable grasp during motion

θ

min

f ∗

si

f ∗

si

f ∗

si f ∗ sinom

f ∗

si + fmi − eT ni(f ∗ si + fmi)eni

eT

ni(f ∗ si + fmi)

≤ µd

such that

Explicit solution in planar two point contacts
Prediction of fm : approximate with projected

measured force

Limiting case: fs à ∞, µd limited by grasp

location

SLIDE 18

Slip Prevention Control

18

If no feasible solution:

θ

f ∗

si + fmi − eT ni(f ∗ si + fmi)eni

eT

ni(f ∗ si + fmi)

≤ µd

Transport: Slow down object

motion

Berthing: Reduce berthing

contact force

SLIDE 19

Slip Prevention Control Experiments

19

7 7 7 7

TABLE I: Comparison of friction threshold violation.

No Compensation (%) Threshold |µ| > µd |µ| > 2µd |µ| > 3µd Case 1, Mean 42.7 16.6 12.2 Case 1, StdDev 2.2 0.5 2.5 Case 2, Mean 43.6 17.1 6.8 Case 2, StdDev 0.6 1.0 0.6 Compensation (%) Threshold |µ| > µd |µ| > 2µd |µ| > 3µd Case 1, Mean 18.7 2.2 0.4 Case 1, StdDev 1.2 0.6 0.3 Case 2, Mean 26.4 5.5 1.3 Case 2, StdDev 1.2 0.4 0.2

Left Right

SLIDE 20

Slip Prevention Control Experiments: Berthing

20

Left Right

SLIDE 21

Slip Prevention Control Experiments: Video

21

SLIDE 22

Space Robot Arm Simulation

22

Based on current data
Both Newton-Euler and

Lagrange-Euler in MATLAB

Joint Friction (linear region to

avoid discontinuity)

Joint Level Control (based
n existing controller design)
Contact Dynamics (simple

switching of dynamics)

Gearing/joint flexibility effect

to be added M(q)¨ q + C(q, ˙ q) ˙ q + F( ˙ q) + G(q) = τ − JT (q)f

SLIDE 23

Future Work

23

Flexible Dynamics Suppression

(Input Shaping)

Docking and grasping

with Marman bands

Space arm simulation (joint

flexibility, grasp stability, docking)

Multiple flying robot grasping /

transport/docking

Flexible material modeling and

handling

SLIDE 24

Soft Material Simulation

24

Simula.on parameters: minimize distance from measured point cloud

Known grasp loca.ons
Blue: measured point cloud
Green: cloth es.ma.on

Best fit Too tight fit Loose fit

Bullet

Posi.on-based dynamics
Approxima.on of physical interac.ons
Fast computa.on
Solu.on for link distance aNer external

forces applied (pulling, gravity, etc.)

SLIDE 25

Other Issues

25

Effect of non-real-time implementation for sensor feedback loop Beyond bimanual manipulation:

SLIDE 26

Acknowledgment

26

Thanks to NASA Goddard for supporting this research and Brian Roberts, Craig Carignan, and Billy Gallagher for guidance and support. Thanks to Center for Automation Technologies and Systems (CATS) at RPI (supported by New York State Office of Science, Technology and Innovation, NYSTAR), and staff and students (Glenn Saunders, Ken Myers, Dan Kruse) for equipment, facilities and technical support.