Space Robotics: Dynamics and Control

290 Pages·1993·13.97 MB·English

by  Robert E. Lindberg, Richard W. Longman (auth.), Yangsheng Xu, Takeo Kanade (eds.)

SPACE ROBOTICS:
DYNAMICS AND CONTROL
THE KLUWER INTERNATIONAL SERIES
IN ENGINEERING AND COMPUTER SCIENCE
ROBOTICS: VISION, MANIPULATION AND SENSORS
Consulting Editor: Tak.eo Kanade
OBJECT RECOGNITION USING VISION AND TOUCH, P. Allen
ISBN: 0-89838-245-9
INTEGRATION, COORDINATION AND CONTROL OF MULTI-SENSOR ROBOT
SYSTEMS, H.F. Durrant-Whyte
ISBN: 0-89838-247-5
BAYESIAN MODELING OF UNCERTAINTY IN LOW·LEVEL VISION, R. Szeliski
ISBN 0-7923-9039-3
VISION AND NA VIGATION: THE eMU NA VLAB, C. Thorpe (editor)
ISBN 0-7923-9068-7
TASK·DIRECTED SENSOR FUSION AND PLANNING: A Computational Approach,
G. D. Hager
ISBN: 0-7923-9108-X
COMPUTER ANALYSIS OF VISUAL TEXTURES, F. Tomita and S. Tsuji
ISBN: 0-7923-9114-4
DATA FUSION FOR SENSORY INFORMATION PROCESSING SYSTEMS, J. Clark
and A. Yuille
ISBN: 0-7923-9120-9
PARALLEL ARCHITECTURES AND PARALLEL ALGORITHMS FOR INTEGRATED
VISION SYSTEMS, A.N. Choudhary, J. H. Patel
ISBN: 0-7923-9078-4
ROBOT MOTION PLANNING, J. Latombe
ISBN: 0-7923-9129-2
DYNAMIC ANALYSIS OF ROBOT MANIPULATORS: A Cartesian Tensor Approach,
C.A Balafoutis, R.V. Patel
ISBN: 07923-9145-4
PERTURBATION TECHNIQUES FOR FLEXIBLE MANIPULATORS: A. Fraser and
R. W. Daniel
ISBN: 0-7923-9162-4
COMPUTER AIDED MECHANICAL ASSEMBLY PLANNING: L. Homen de Mello and
S.Lee
ISBN: 0-7923-9205-1
INTELLIGENT ROBOTIC SYSTEMS FOR SPACE EXPLORATION: Alan A. Desrochers
ISBN: 0-7923-9197-7
MEASUREMENT OF IMAGE VELOCITY: David J. Fleet
ISBN: 0-7923-9198-5
DIRECTED SONAR SENSING FOR MOBILE ROBOT NA VIGATION: John J. Leonard,
Hugh F. Durrant-Whyte
ISBN: 0-7923-9242-6
A GENERAL MODEL OF LEGGED LOCOMOTION ON NATURAL TERRAIN: David
J.Manko
ISBN: 0-7923-9247-7
INTELLIGENT ROBOTIC SYSTEMS: THEORY, DESIGN AND APPLICATIONS K.
Valavanis, G. Saridis
ISBN: 0-7923-9250-7
QUALITATIVE MOTION UNDERSTANDING: W. Burger, B. Bhanu
ISBN: 0-7923-9251-5
DIRECTED SONAR SENSING FOR MOBILE ROBOT NAVIGATION: J.J. Leonard,
H.F. Durrant-Whyte
ISBN: 0-7923-9242-6
SPACE ROBOTICS:
DYNAMICS AND CONTROL
Edited by:
Yangsheng Xu
Takeo Kanade
Carnegie Mellon University
.....
"
SPRINGER SCIENCE+BUSINESS MEDIA, LLC
Library of Congress Cataloging-in-Publication Data
Space robotics : dynamics and control / edited by Yangsheng Xu, Takeo
Kanade.
p. cm. -- (fhc Kluwer international series in engineering and computer
science ; SECS 188)
Includes bibliographical references and index.
ISBN 978-1-4613-6595-2 ISBN 978-1-4615-3588-1 (eBook)
DOI 10.1007/978-1-4615-3588-1
1. Space stations--Automation. 2. Robotics. 1. Xu, Yangsheng. II.
Kanade, Takeo. III. Series.
TL797.S6445 1992 92-28350
629.47--dc20 CIP
Cover illustration courtesy of Canadian Space Agency.
Painting by Paul Fjeld, Canadian artist. .
Copyright © 1993 by Springer Science+Business Media New York
Originally published by Kluwer Academic Publishers in 1993
Softcover reprint ofthe hardcover Ist edition 1993
AII rights reserved. No part of this publication may be reproduced, stored in a retrieval
system or transmitted in any form or by any means, mechanical, photo-copying, record ing,
or otherwise, without the prior written permission of the publisher, Springer Science
+Business Media. LLC
Printed on acid-free paper.
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. vii
1. Kinematic and Dynamic Properties of an Elbow Manip
ulator Mounted on a Satellite Robert E. Lindberg,
Richard W. Longman and Michael F. Zedd ... . . . . . . . . . . . .. 1
2. The Kinetics and Workspace of a Satellite-Mounted
Robot Richard W. Longman ......................... 27
3. On the Dynamics of Space Manipulators Using the Virtual
Manipulator, with Applications to Path Planning
Z. Vafa and S. Dubowsky ........................... 45
4. Dynamic Singularities in Free-noating Space Manipulators
Evangelos Papadopoulos and Steven Dubowsky .... . . . . . . .. 77
5. Nonholonomic Motion Planning of Free-Flying Space
Robots via a Bi-Directional Approach Yoshihiko
Nakamura and Ranjan Mukherjee . . . . . . . . . . . . . . . . . . . .. 101
6. Attitude Coutrol of Space Platform/Manipulator System
Using Internal Motion
C. Fernandes, L. Gurvits and Z.x. Li .................. 131
7. Control of Space Manipulators with Generalized Jacobian
Matrix Kazuya Yoshida and Yoji Umetani . . . . . . . . . . . . .. 165
8. Sensory Feedback Control for Space Manipulators
Yasuhiro Masutani, Fumio Miyazaki, and Suguru Arimoto ... 205
9. Adaptive Control of Space Robot System with an
Attitude Controlled Base Yangsheng Xu, Heung-Yeung Shum,
Ju-Jang Lee, and Takeo Kanade .. . . . . . . . . . . . . . . . . . . .. 229
10. Experiments in Autonomous Navigation and Control of a
Multi-Manipulator, Free-Flying Space Robot
Marc A. Ullman and Robert H. Cannon, Jr. ......... . . . .. 269
Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 285
Preface
Robotic technology offers two potential benefits for future space exploration. One
benefit is minimizing the risk that astronauts face. In the inhospitable environment
of space, humans must work under extreme temperature, glare, and possibly high
levels of radiation. The other benefit is increasing their productivity. Extra
vehicular activity (BV A) consumes considerable time and can require a load handling
capacity and dexterity exceeding human capability.
Realizing the benefits of robotic technology in space will require solving several
problems. Some of them are unique and now becoming active research topics. One
of the most important research areas is dynamics, control, motion and planning for
space robots by considering the dynamic interaction between the robot and the base
(space station, space shuttle, or satellite). Due to the dynamic interaction, the
motion of space robots can alter the base trajectory. On the other hand, the robot
end-effector may miss the desired target due to the motion of the base. This mutual
dependence severely affects the performance of both the robot and the base,
especially when the mass and the moment of inertia of the robot and payload are not
negligible in comparison to the base. any inefficiency in the planning and control
can considerably risk the success of the space mission.
For this book, we collected recent research papers on fundamental problems in
dynamics and control of space robots, especially issues relevant to dynamic
base/robot interaction. The authors are pioneers in theoretical analysis and
experimental system development of space robot technology. Our goal is to provide
a reference work for researchers and graduate students in robotics, mechanics,
control, and astronautical science. We organized the papers under three problem
areas: dynamics problems, nonholonomic nature problems, and control problems.
In the first part, four papers extensively investigate dynamic problems of
interaction between a space robot and its base. Lindberg, Longman, and Zedd
address several issues related to dynamics and kinematics of a space robot system
when the base in controlled in orientation but free in translation, and when the base
is free in both orientation and translation. When the base is free in both orientation
and translation, the kinematic mapping between joint space and inertial space
depends on not only dynamic parameters, such as mass and inertia, but also the
motion history, i.e., the path the robot has followed. In the second paper, Longman
discusses the kinematic relationship in joint space and inertial space, calledforward
and inverse kinetics problems, and the workspace of a space robot. The paper by
Vafa and Dubowsky introduces the concept of Virtual Manipulator to represent the
dynamics of a space robot system. The Virtual manipulator concept makes it
possible to reproduce the kinematic behavior of a space robot by the kinematics of
a modified fixed-base robot. They applied this concept to plan robot motions that
minimize disturbances to the spacecraft. The virtual manipulator concept is used by
Papadopoulos and Dubowsky to study the singularity problem of space robots.
The conservation of angular momentum imposes non-integrable velocity constraints
that result in the nonholonomic nature of space robots. We selected two papers, one
by Nakamura and Mukherjee representing early work on this topic, and the other by
Fernandes, Gurvits, and Li representing a recent effort toward optimal attitude
control using internal motion.
viii
The final papers address control problems of space robots. Yoshida and Umetani
propose a resolved rate and acceleration control based on the Generalized Jacohin
Matrix of a space robot, and then discuss the use of the control scheme i the capture
of stationary and moving objects. Masutani, Miyazaki, and Arimoto address feed
back control problems of a space robot system. The paper by Xu, Shum, Lee, and
Kanade presents an adaptive control scheme for a space robot system in the presence
of uncertainty. The paper also discusses the parameterization nonlinearity in terms
of the dynamic parameters of a space robot in inertia space. The last paper by
Ullman and Cannon provides an experimental study in autonomous navigation and
control of multiple free-flying space robots.
Yangshmg Xu
Takeo Kanade
SPACE ROBOTICS:
DYNAMICS AND CONTROL
Kinematic and Dynamic
Properties of an Elbow
Manipulator Mounted on a
Satellite
Robert E. Lindbergl, Richard W. Longman2, and Michael F. Zedd3
Abstract
The discussion in this paper is intended as an introduction to several topics treated in
various forms or extensions in the other papers in this issue. Many applications of robots
in space require the robot to manipulate a load mass that is not negligible compared to
the satellite on which the robot is mounted. In such cases, the robot motion disturbs the
position and attitude of the satellite, and this in turn disturbs the robot end-effector posi
tion. This implies that the robot joint angles that would normally be commanded to pro
duce a prescribed robot end-effector position and orientation will result in missing the
target. In this paper an overview is given of certain engineering problems arising from
this phenomenon. The robot end-effector positioning problem is discussed for the two
cases of the satellite attitude control system either off or on, and computation of the robot
motion disturbances to the satellite is discussed. Shuttle Remote Manipulator System
based examples are given.
Introduction
When a robot manipulator is mounted on a satellite, there is an interaction be
tween the robot dynamics and the dynamics of the satellite. This raises several
questions: What happens to the problem of commanding the robot to go to some
desired position, when the base on which the robot is mounted-i.e. the satel
lite-does not stay in a fixed inertial position? And, conversely, what happens to
Iprogram Manager, Orbital Sciences Corporation, Fairfax, VA 22033; formerly, Head, Concept
Development Branch, Spacecraft Engineering Department, Naval Research Laboratory, Wash
ington, D.C. 20375.
2Expert Consultant, Naval Research Laboratory, Washington, D.C. 20375; also, Professor of Me
chanical Engineering, Columbia University, New York, NY 10027.
JAerospace Engineer, Spacecraft Engineering Department, Naval Research Laboratory, Wash
ington, D.C. 20375.
The Journal of the Astronautical Sciences, Vol. 38, No.4, 1990, pp. 397-421. Copyright © 1990
by the America Astronautical Society Inc., reproduced here with kind permission of the America
Astronautical Society, Inc.
2
the satellite position and orientation as a result of the robot motion? These ques
tions are important whenever the mass or the inertia of the object manipulated
by the robot is not totally negligible compared to the mass or the inertia of the
spacecraft.
Concerning the first question, two different situations can apply. In one case,
the satellite is totally free to both translate and rotate as a result of the robot mo
tion disturbances. In the other case, there is an active attitude control system
that maintains the satellite attitude relative to inertial space by applying appro
priate corrective torques, but the satellite center of mass is still free to translate
in response to robot force disturbances. One can of course imagine still a third
alternative in which the satellite includes some kind of active position control
system, that can make the satellite into an inertial platform for mounting the
robot by cancelling robot force disturbances as well as torque disturbances. This
would be an unusual satellite, but such systems may be necessary to maintain the
micro-gravity environment in multi-user satellites.
The main purpose for this paper appearing in this special issue on Robotics in
Space is to serve as an introduction to many of the salient features of satellite
mounted robot problems. The main topics discussed are listed below, and the
summary at the end of the paper indicates how these topics are treated or ex
tended in other papers in the issue:
1. Robots mounted on satellites that are free to both translate and rotate are
discussed and illustrated by an illuminating example. It is shown that the
position of the robot end-effector is no longer just a function of the present
robot joint angles, but rather a function of the whole history of the joint
angles. This causes the usual forward and inverse kinematics problems in
robotics to become problems with a totally different character. Reference
[1], which is in this issue, coins the terms forward kinetics and inverse ki
netics for these new problems.
2. When the satellite attitude is maintained by an attitude control system dur
ing a robot maneuver, or when the attitude is reacquired at the end of a
robot maneuver, the forward and inverse kinematics problems are still
kinematics problems in which the end-effector position is purely a function
of the final robot joint angles. However, a new kinematics must be used to
account for the translational motion of the satellite, or robot base, associ
ated with the motion of the centers of mass of each robot link and the robot
load mass. The Remote Manipulator System mounted on the Shuttle is used
as a model for developing this space-based kinematics. The development
parallels that of the earlier paper [2] dealing with a polar coordinate robot,
and was presented in full detail in the conference version of this paper
which appeared as reference [3].
3. The need to use more than one form of kinematics on orbit is discussed. In
some situations, one uses the standard kinematics, and in other situations
one uses the new space-based kinematics with various required reinitializa
tions during maneuvers. Use of matching conditions to convert between
different sets of kinematic equations is also summarized.