Abstract
〈Vol.1 No.6(2008.6)〉
Titles
▲
■ Stability of Dynamical Systems with Discontinuous Motions: Beyond Classical
Lyapunov Stability Results
University of Notre Dame・Anthony N. MICHEL,and St. Cloud State University・Ling
HOU
In this paper we present a stability theory for discontinuous dynamical
systems (DDS): continuous-time systems whose motions are not necessarily
continuous with respect to time. We show that this theory is not only applicable
in the analysis of DDS, but also in the analysis of continuous dynamical
systems (continuous-time systems whose motions are continuous with respect
to time), discrete-time dynamical systems (systems whose motions are defined
at discrete points in time) and hybrid dynamical systems (HDS) (systems
whose descriptions involve simultaneously continuoustime and discrete-time).
We show that the stability results for DDS are in general less conservative
than the corresponding well-known classical Lyapunov results for continuous
dynamical systems and discrete-time dynamical systems. Although the DDS
stability results are applicable to general dynamical systems defined on
metric spaces (divorced from any kind of description by differential equations,
or any other kinds of equations), we confine ourselves to finite-dimensional
dynamical systems defined by ordinary differential equations and difference
equations, to make this paper as widely accessible as possible. We present
only sample results, namely, results for uniform asymptotic stability in
the large.
▲
■ Automatic Sleep Stage Determination by Multi-Valued Decision Making
Based on Conditional Probability with Optimal Parameters
Saga University/East China University of Science and Technology・Bei WANG,,
Saga University・Takenao SUG,East China University of Science and Technology・Xingyu WANG,
and Saga University・Masatoshi NAKAMURA
Data for human sleep study may be affected by internal and external influences.
The recorded sleep data contains complex and stochastic factors, which
increase the difficulties for the computerized sleep stage determination
techniques to be applied for clinical practice. The aim of this study is
to develop an automatic sleep stage determination system which is optimized
for variable sleep data. The main methodology includes two modules: expert
knowledge database construction and automatic sleep stage determination.
Visual inspection by a qualified clinician is utilized to obtain the probability
density function of parameters during the learning process of expert knowledge
database construction. Parameter selection is introduced in order to make
the algorithm flexible. Automatic sleep stage determination is manipulated
based on conditional probability. The result showed close agreement comparing
with the visual inspection by clinician. The developed system can meet
the customized requirements in hospitals and institutions.
▲
■ Pole Constraints of Reference Models in 2-DOF Servo System Design for
Non-Minimum Phase Systems
Osaka University・Toru Asai,Kumamoto University・Hiroshi Okajima amd
Dresdner Kleinwort (Japan) Limited・Chee Leong Ooi
This paper is concerned with the analysis of pole constraints in servo
system design for non-minimum phase (NMP) systems. We first characterize
the achievable closed-loop system for a SISO plant. For simplicity, we
assume that the plant has only one NMP zero. Based on the characterization
and the tracking condition, we show for some combinations of degree and
relative degree of the closed-loop system that the admissible location
of poles is restricted. For these cases,
we provide a quantitative measure for the limitation. We also provide its
concrete formula for some specific cases.
▲
■ Robust Stability Analysis of Uncertain Interconnections
Kyoto University・Kiyotsugu TAKABA
This paper considers robust stability of an interconnection of a linear
time-invariant differential nominal system and passive uncertainties in
the behavioral framework. A generalized version of the well-known passivity
theorem is formulated by using quadratic differential forms. Based on the
generalized passivity theorem, it is proved that, if the nominal system
is $\Phi$-passive, the interconnection is robustly stable against strictly
$(-\Phi)$-passive uncertainty. Moreover, we show that the $\Phi$-passivity
of the nominal system is a necessary and sufficient condition for robust
stability with regularity constraint of the uncertain interconnection.
▲
■ Remote hybrid controller with adaptive environment observer
Nara Institute of Science and Technology・
Yuki Kakizoe, Hisakazu Nakamura and Hirokazu Nishitani
We propose a method of remote bilateral-control of master-slave manipulators
with an environment observer. The adaptive environment observer estimates
characteristics of the environment in a remote place. Thus, the system
predicts a virtual environment that is generated by an actual
environment. The proposed system exhibits good operatability and tracking
when the slave system presses a compliant wall under the condition of information
transmission delay. We confirm the effectiveness and safety of the proposed
system by experiments.
▲
■ Sensor-less Interaction Force Control Based on B-Spline Function for
Human-Robot Systems
Nagaoka University of Technology・Chowarit Mitsantisuk,Keio University・Seiichiro
Katsura,
and Nagaoka University of Technology・Kiyoshi Ohishi
In this paper, to provide precise force sensation of human operator, the
twin direct-drive motor system with wire rope mechanism has been developed.
The human-robot interaction force control is independently integrated in
acceleration dimension by realizing the dual disturbance observer based
on modal space design.
In the common mode, it is utilized for control of vibration suppression
and wire rope tension. In the differential mode, the purity of human external
force with compensation of friction force is obtained. This mode is useful
for control the interaction force of human. Furthermore, the human-robot
system that has the ability of support of human interaction force is also
proposed. The interaction force generation based on B-spline function is
applied to automatically adjust the smooth force command corresponding
to the adaptive parameters.
▲ ■ Time synchronized wireless sensor network for vibration measurement
Shibaura Institute of Technology・Yutaka Uchimura,
Kajima Corporation・Tadashi Nasu, Motoichi Takahashi
In this paper, we introduce a newly developed time synchronized wireless
sensor network system. The system employs IEEE 802.11 standard based TSF
counter and sends the measured data with the counter value. TSF based synchronization
enables consistency on common clock among different wireless nodes. We
consider the scale effect on the synchronization accuracy and the effect
is evaluated by stochastic analysis and simulation studies. A new wireless
sensing system was
developed and the hardware and software specifications are shown. The experiments
were conducted in a reinforced concrete building and results showed that
the system performed more than sufficiently.
▲
■ A Branch and Bound Method to the Continuous Time Model Elevator System
with Full Information
Tsinghua University/Boston University・Zhen SHEN,and Tsinghua University・Qianchuan ZHAO
We give a new Branch and Bound method for the scheduling of the group elevator
system with full information. Full information means that not only the
parameters of the elevator systems but also the arrival time, origins and
destinations of all the passengers who are to be served are known beforehand.
The performance obtained by solving the full information problem is the
best performance that the elevator scheduling algorithm can achieve and
then can be used to measure how good an elevator scheduling algorithm is.
Our method can handle the continuous time event and is based on the concept
of “trip”, which refers to the movement of the car without changing the
direction and with at least one passenger being served.
▲ ■ On the Exactness Proof of (D, G) Scaling
Kyoto University・Yoshio EBIHARA
The goal of this paper is to provide an elementary proof for the exactness
of the 0(D,G) scaling.
The (D,G) scaling has vast application area around control theory, optimization
and signal processing. This is because, by applying the (D,G) scaling,
we can convert inequality conditions depending on an uncertain parameter
to linear matrix inequalities (LMIs) in an exact fashion. However, its
exactness proof is tough, and this stems from the fact that the proof requires
an involved matrix formula in addition to the standard Lagrange duality
theory. To streamline the proof, in the present paper, we clarify that
the involved matrix formula is closely related to a norm preserving dilation
under structural constraints. By providing an elementary proof for the
norm preserving dilation, it follows that basic results such as Schur complement
and congruence transformation in conjunction with the Lagrange duality
theory are enough to complete a self-contained exactness proof.
|
