1
Theoretical and experimental evaluation of a
2-Channel bilateral Force Reflection teleoperation
system
A. Frisoli , E. Sotgiu, D. Checcacci, F. Simoncini, S. Marcheschi, C. A. Avizzano, M. Bergamasco
PERCRO - Scuola Superiore Sant’Anna
Pisa - Italy
Phone: +39050883080, Fax: +39050883333, Email: antony@sssup.it, sotgiu@sssup.it, carlo@sssup.it,
bergamasco@sssup.it
Abstract—This paper presents the theoretical and experimental
evaluation of a 2-Channel bilateral Force Reflection
teleoperation system.
The teleoperation system is a master-slave bilateral system
based on two parallel kinematics robots with three degree of
freedom, which has been devised for the validation of robotics
assisted surgery procedures. Both master and slave systems
present an isomorphic kinematics which allow only translational
motion of the end-effector.
The aim of the present work is to evaluate two different
implementations of a Force Reflection control architectures: a
Position Error Based Force Reflection Control and a Direct
Force Reflection Control. The two architectures are compared
in terms of performance, i.e. stability and transparency, on the
basis of both simulations and experimental data. Some novel
theoretical results are presented, which permit a direct evaluation
of performance for both the two architectures.
I. INTRODUCTION
The rendering to the human operator of the sensation
of physical interaction with a remote environment can be
achieved by teleoperator systems. Recently studies concerning
the properties of teleoperator systems have focused on applications
in computer assisted surgery [Flemmer and Wikander,
2003], where the quality of the force-feedback to the operator
becomes more crucial than in other applications.
Passivity and classical stability analysis represent the two
common approaches for the analysis of teleoperation systems.
Analysis of absolute stability [Llewellyn, 1952], and more
generally passivity analysis, represents a correct theoretical
framework for the analysis of stability of teleoperator systems,
but unfortunately it can result to be too restrictive in some
cases, and provide only global information on the system
behavior. Classical control analysis, in absence of transmission
delays, can provide better insight on the sensitivity/robustness
of the control to the variation of some parameters [Daniel and
McAree, 1998], by exploiting graphic representations such as
the root locus.
Several architectures can be implemented [Hashtrudi-Zaad
and Salcudean, 2001],[Arcara and Melchiorri, 2002] according
to the desired performance of the system. For the scope of
the present paper, two implementations of a Force-Position
(F-P) scheme (according to the definition by [Hashtrudi-Zaad
and Salcudean, 2001]) will be theoretically and experimentally
compared, on the basis of their performance in terms of
both stability and passivity. The F-P scheme (see Figure I)
presupposes that the Slave system acts as a force sending
device, while the Master as a position sending device; it is
commonly known in literature also as the Force Reflection
(FR) scheme [Arcara and Melchiorri, 2002].
In the first arrangement of the FR scheme the force fedback
to the master is the impedance force Fcs generated by the
slave controller [Arcara and Melchiorri, 2002], as shown in
Figure 1(a). This scheme is often referred to as the Position
Error Based Force Reflection Control. This architecture, later
on indicated with the short notation Fcs-P, is usually adopted
when the interaction force Fe at slave side is not available and
it is an alternative to the direct measure of Fe.
A second implementation of the FR scheme is the Direct
Force Reflection Control [Daniel and McAree, 1998], later on
called Fe-P (see Figure 1(b)), and it presupposes instead that
the measure of interaction force Fe is available at the slave side
and is directly fedback to the master. A hybrid approach is the
Shared Compliance Control [Arcara and Melchiorri, 2002], but
it will not be analyzed in the present study.
(a) Fcs-P scheme. (b) Fe-P scheme.
Fig. 1. 2-Channel Teleoperation Systems.
It is interesting to compare quantitatively the performance of
.....