Abstract
〈Vol.3 No.2(2010.3)〉
Titles
[Contributed Papers]
▲
■ Removing Shift Variant Blur without Geometric Coordinate Transformation
Ritsumeikan University・Wikky Fawwaz Al MAKI
and Sueo SUGIMOTO
In this paper, we present an image deconvolution algorithm for shift variant
blur caused by a relative circular motion between the imaging equipment
and the object during the exposure time. The two-dimensional shift variant
blurred image caused by the accelerated circular motion is modeled as an
output of a series of one-dimensional linear convolutions that expresses
the degradation process. To simulate the accelerated circular motion blurred
image, the midpoint circle algorithm is employed. In the image decovolution
process, for the given center of the circular motion, the algorithm is
used to obtain circular paths at which the degradation process happens.
The shift invariant deconvolution method is applied to each circular path.
After the estimated true image is obtained, median filtering is applied
to solve the some empty pixels. Simulation experiments confirm the capability
of the proposed method.
▲
■ Automated Cell-Cutting for Cell Cloning
AIST・Akihiko ICHIKAWA, Tamio TANIKAWA,
NILGS ・Kazutsugu MATSUKAWA, SeiyaTAKAHASHI,
and AIST・Kohtaro OHBA
We develop an automated cell-cutting technique for cell cloning. Animal cells softened by the cytochalasin treatment are injected into a microfluidic chip. The microfluidic chip containe two orthogonal channels: one microchannel is wide, used to transport cells, and generates the cutting flow; the other is thin and used for aspiration, fixing, and stretching of the cell. The injected cell is aspirated and stretched thin by transporting it to the thin microchannel. Simultaneously, the volumes of the cell before and after aspiration are calculated; the volumes are used to calculate the fluid flow required to aspirate half the volume of the cell into the thin microchannel. Finally, we apply a high-speed flow in the orthogonal microchannel to bisect the cell. This paper reports the cutting process, the cutting system, and the results of the experiment.
▲
■ Construction of a Cr3C2-C Peritectic Point Cell for Thermocouple Calibration
NMIJ/AIST・Hideki OGURA,
LNE, Paris・Thierry DEUZE, Ronan MORICE,
LNE, Trappes ・Pascal RIDOUX, and Jean-Remy FILTZ
The melting points of Cr3C2-C peritectic (1826 °C) and Cr7C3-Cr3C2 eutectic
(1742 °C) alloys as materials for high-temperature fixed point cells are
investigated for use in thermocouple calibration. Pretests are performed
to establish a suitable procedure for constructing contact thermometry
cells based on such chromium-carbon mixtures. Two cells are constructed
following two different possible procedures. The above two melting points
are successfully observed for one of these cells using tungsten-rhenium
alloy thermocouples.
▲
■ Current Sensor based Home Appliance and State of Appliance Recognition
Tottori University・Takeshi SAITOH, Tomoyuki OSAKI,
Ryosuke KONISHI, and Kazunori SUGAHARA
This paper presents the current sensor-based home appliance and its state
recognition method for intelligent outlets. Our system has three main functions:
remote control, monitor, and power supply schedule management. This research
focuses particular on the monitor function. To recognize the appliance
and the state of the appliance, we extract ten features based on a measured
current signal. In the experiment, we gather a number of signals with various
appliances, and find that three features $I_{rms}$, $I_{avg}$, and $I_{peak}$
yield valid recognition results of 84.3%, 86.4%, and 90.3% for classifying
the state of the appliance into three categories. Moreover, sufficient
recognition rates of 97.4%, 97.7%, and 99.0% are obtained by consideration
of three candidates.
▲
■ Seamless Transmission Between Single-mode Optical Fibers Using Free
Space Optics System
NTT Access Network Service Systems Laboratories・
Koichi YOSHIDA and Takeshi TSUJIMURA
This paper presents a free space optics system installed between two single-mode optical fibers (SMFs). The result is as if the two SMFs were seamlessly connected without the need for any photoelectric devices. Misalignments between the two SMFs caused by disturbances are actively compensated for by introducing a laser beam controller that incorporates an opto-mechatronic mechanism with four degrees of freedom. Experiments using a prototype are conducted to verify the effectiveness of the proposed FSO system for initial beam acquisition and beam tracking when there is vibration disturbance.
▲
■ Robust lp stability and robust l^infinity stabilization over a rate-limited
digital channel
Kyoto University・Yumiko ISHIDO and Kiyotsugu TAKABA
This paper is concerned with robust stabilization of an uncertain system
over a rate-limited digital channel from the viewpoint of input-output
stability. Because of the quantization in the communication channel, it
is impossible to apply the traditional small gain theorem to establish
robust stability of the feedback system. To overcome this difficulty, we
introduce a new notion of small $\ell^p$ signal $\ell^p$ stability, and
derive a sufficient robust stability condition against $\ell^p$ gain-bounded
uncertainty based on this notion. Furthermore, for the case of $p=\infty$,
a sufficient data rate for the existence of a robustly stabilizing encoder-controller
pair is explicitly given.
▲ ■ A Study on Stability of Optimal Regulators under Online Gain Adjustment
Kyoto University・Takayuki INUI and Tomomichi HAGIWARA
The objective of this paper is to show a general framework for online adjustment
methods of the state feedback gains for multi-input multi-output optimal
regulators via the quadratic stability arguments. We first give a region
of state feedback gains containing the nominal optimal gain such that quadratic
stability is ensured for all time-invariant state feedback gains in the
region. We then show that under some mild conditions, quadratic stability
is retained for arbitrary time-varying adjustment of the state feedback
gain within this region via the concept of strong uniform complete observability.
It is then illustrated how the result can be utilized as a general framework
for constructing heuristic online gain adjustment laws, and the usefulness
of the result is confirmed with a numerical example.
▲
■ Genetic Network Programming with Reconstructed Individuals
Waseda University・Fengming YE, Shingo MABU, Lutao WANG,
Shinji ETO, and Kotaro HIRASAWA
A lot of research on evolutionary computation has been done and some significant classical methods such as Genetic Algorithm (GA), Genetic Programming (GP), Evolutionary Programming (EP) and Evolution Strategies (ES), have been studied. Recently, a new approach named Genetic Network Programming (GNP) has been proposed. GNP can evolve itself and find the optimal solutions. It is based on the idea of Genetic Algorithm and uses the data structure of directed graphs. Many papers have demonstrated that GNP can deal with complex problems in the dynamic environments very efficiently and effectively. As a result, recently, GNP is getting more and more attentions and is used in many different areas such as data mining, extracting trading rules of stock markets, elevator supervised control systems, etc and GNP has obtained some outstanding results. Focusing on the GNP's distinguished expression ability of the graph structure, this paper proposes a method named Genetic Network Programming with Reconstructed Individuals (GNP-RI). The aim of GNP-RI is to balance the exploitation and exploration of GNP, that is, to strengthen the exploitation ability by using the exploited information extensively during the evolution process of GNP and finally obtain the better performances than that of GNP. In the proposed method, the worst individuals are reconstructed and enhanced by the elite information before undergoing genetic operations (mutation and crossover). The enhancement of worst individuals mimics the maturing phenomenon in nature, where bad individuals can become smarter after receiving a good education. In this paper, GNP-RI has been applied to the tile-world problem which is an excellent bench mark for evaluating the proposed architecture. The performance of GNP-RI is compared with that of the conventional GNP. The simulation results show some advantages of GNP-RI over the conventional GNPs demonstrating its superiority.
▲
■ A Weight Compensation Mechanism with a Non-Circular Pulley and a Sping:
Application to a Parallel Four-bar Linkage Arm
Tokyo Institute of Technology・Gen ENDO, Hiroya YAMADA,
Canon Inc.・Akira YAJIMA, Masaru OGATA,
and Tokyo Institute of Technology ・Shigeo HIROSE
We propose a new weight compensation mechanism with a non-circular pulley
and a spring. We show the basic principle and a numerical design method
to derive the shape of the non-circular pulley. After demonstration of
the weight compensation for an inverted pendulum system, we extend the
same mechanism to a 2 degrees of freedom (DOF) parallel four-bar linkage
system, analyzing the required torques with transposed Jacobian matrices.
Finally, we develop a 3-DOF manipulator with relatively small output actuator
and verifies that the weight compensation mechanism significantly contributes
to decrease static torques to keep the same posture within manipulator's
work space, which is also effective in a dynamic movement.
▲
■ Parameterization of the Output with respect to Relative Degree
Tokyo Institute of Technology・Kazuma SEKIGUCHI, Mitsuji SAMPEI,
and Sasebo National College of Technology・Shigeki NAKAURA
The partial feedback linearization is capable of expressing a complicated
nonlinear system as a simple form to control. To get a proper linearized
subsystem, it is most important to select a proper coordinate transformation.
This paper plays a fundamental role in selecting the proper coordinate
by revealing the system structure with respect to the relative degree.
A relative degree is a key concept to treat the partial feedback linearization,
especially, the input-output linearization for an input affine system.
Using special functions characterized in the relative degree, all functions
are parameterized and classified according to their relative degrees. Applying
the proposed parameterization to a coordinate transformation, the system
structure with respect to relative degrees is revealed. Moreover using
the input-output linearization with parameterized outputs, it is also shown
that the internal dynamics of the partially linearized system is parameterized.
|
