IEEJ Transactions on Electronics, Information and Systems
Online ISSN : 1348-8155
Print ISSN : 0385-4221
ISSN-L : 0385-4221
Volume 124, Issue 9
Displaying 1-39 of 39 articles from this issue
Special Issue on “Sensing and Imaging of Biomedical Information”
Special Issue Paper
<Biomedical Engineering>
  • Akinori Ueno, Yoichi Furusawa, Hiroshi Hoshino, Yoji Ishiyama
    2004 Volume 124 Issue 9 Pages 1664-1671
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    This article reports on a novel technique for detecting electrocardiogram (ECG) at a condition where thin cloth is interpolated between sensing electrodes and the skin to which the electrodes are attached. The technique is based upon capacitive coupling composed of the electrode, the cloth and the skin, so that the electrode can lead alternating electrocardiographic current through capacitance of the coupling. The technique is also founded on impedance transforming circuit that has extremely high input impedance around 1000GΩ and low output impedance, so as to match high output impedance of the electrode to low input impedance required by subsequent circuitry. A pilot ECG measuring device was manufactured using the technique and experiments showed (1) ECG recordings using the device with silk of 240μm thickness or with cotton of 564μm thickness were quite similar to ECGs recorded from the skin using conventional system, (2) stable ECGs were observed with the silk below 600μm thickness or with the cotton below 1128μm thickness, (3) effects of long-term measurement and perspiration on ECG waveform were negligible. These results prove feasibility of the proposed technique for detecting ECG by electrodes with fabrics.
    Download PDF (6922K)
  • Yuichi Ogawa, Kodo Kawase, Maya Mizuno, Masatsugu Yamashita, Chiko Ota ...
    2004 Volume 124 Issue 9 Pages 1672-1677
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    We constructed a THz transillumination system for water content monitoring, and we succeeded in measuring the moisture level in plants. Our measurement system uses a widely tunable coherent THz parametric oscillator source. As target we chose for this experiment a leaf of Japanese basil. The time variation of the water content in the leaf was monitored in two situations: a leaf freshly cut which is left to dry out, and the leaf of a water stressed plant. We found by real-time measurements that the water content of a cut leaf does not decrease uniformly in time. Also, the response to water stress is delayed by about 5-10 minutes. Furthermore, we demonstrated a moisture measurement using a transillumination THz imaging system. As target we chose for this experiment a leaf of Hedera helix held between two thin plastic sheets. The change of the moisture distribution is clearly visible. These results show that the method described here can be applied to nondestructive and real-time monitoring of water content in plants.
    Download PDF (5011K)
  • Masayoshi Tsubai, Toshihiro Nishimura
    2004 Volume 124 Issue 9 Pages 1678-1684
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    We discuss a new morphological operations using double structuring elements (DSEs) and apply them to edge sharpening of the ultrasound images with speckle. The DSEs, a couple of two different structuring elements, varying with the local characteristics of the processing image, yield an edge-sharpening effect, which general single structuring element does not have. Our method sharpens the boundary without speckle emphasis, overshoot nor undershoot because it is based on geometrical characteristics of the image components.
    Download PDF (250K)
  • Masaki Kawakatsu, Hidenori Ishibashi, Masaharu Adachi, Yoshinori Uchik ...
    2004 Volume 124 Issue 9 Pages 1685-1691
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    In the biomagnetic measurement, the biomagnetic signal is extremely weak compared with environmental magnetic noise. Therefore, it is important to reduce the noise component. There are many noise-reduction studies for MEG using Independent Component Analysis (ICA). The ICA method is expectable to extract and remove noise components from the brain magnetic field measurement data. However, in these researches, each obtained independent components are artificially distinguished to the noise and the signal. We propose a method of distinguishing to the noise and the signal automatically by using the signal subspace method for vector brain magnetic field. By applying this method to a phantom data and Auditory Evoked Field data, it is shown improvement of the signal to noise ratio and estimated accuracy.
    Download PDF (5615K)
  • Tadanori Fukami, Takamasa Shimada, Takao Akatsuka, Yoichi Saito
    2004 Volume 124 Issue 9 Pages 1692-1697
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    In audiometry, ABR (Auditory Brainstem Response) is widely used. However, it shows low accuracy in low frequency band. Meanwhile, AMFR (Amplitude-Modulation-Following Response), the response during hearing an amplitude-modulated tone, has high frequency specificity and is brought to attention. As the first step to clinical application of AMFR, we investigated the activated areas in a brain when the subjects hear SAM tone (Sinusoidally Amplitude-Modulated tone) with both ears. We measured following two signals. One is the difference of BOLD (Blood Oxygenation Level Dependent) signal between hearing SAM tone vs. silence, the other is the difference of BOLD signal between hearing SAM tone vs. unmodulated tone. As a result, in the case of SAM vs. silence, the bilaterally auditory cortex (Broadmann Area 41, 42), the biratelally BA 10, left superior frontal gyrus and right superior temporal gyrus were activated (p<0.0037, uncorrected).
    In the case of SAM vs. unmodulated tone, the bilaterally superior frontal gyrus (BA 6) and precuneus (BA 7), neighboring area including the bilaterally inferior parietal lobule (BA 40), the bilaterally medial frontal gyrus and superior frontal gyrus were activated (p<0.021, uncorrected). Activations of visual perception due to eye-opened state were detected in some parts of activations.
    As a result, we inferred that modulated tone was recognized in the medial frontal gyrus and inferior parietal lobule was the part related to perception of amplitude-modulation.
    Download PDF (168K)
  • Yoshiharu Koya, Mitsuyo Ito, Isao Mizoshiri
    2004 Volume 124 Issue 9 Pages 1698-1703
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    Modeling and analysis of the circulation system enables the characteristic decision of circulation system in the body to be made. So, many models of circulation system have been proposed. But, they are complicated because the models include a lot of elements. Therefore, we proposed a complete circulation model as a lumped electrical circuit, which is comparatively simple.
    In this paper, we examine the effectiveness of the complete circulation model as a lumped electrical circuit. We use normal, angina pectoris, dilated cardiomyopathy and myocardial infarction for evaluation of the ventricular contraction function.
    Download PDF (4069K)
  • Iwao Okutani, Tatsuo Takase
    2004 Volume 124 Issue 9 Pages 1704-1711
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    Interactions between regional landscape and sound are investigated in terms of subjects’ brain wave properties, i.e., α wave power spectrum and frequency fluctuation index of brain wave. The landscape samples consist of sea, buildings, crowds, trees, flowers, streamlet, vehicular traffic flow and residential area while the sound samples are made up of crowds sound, insect chirpings, rustling sound of leaves, wave sound, stream murmurings, bird chirpings and supersonic. It is revealed that (1) most of the best combinations of landscape and sound include natural landscape and sound whereas the worst combinations are associated with artificial environments, (2) best(worst) combinations determined via fluctuation index are composed of the landscape and sound which are evaluated as bests(worsts) at landscape only or sound only stimulus tests, (3) based on theα wave power spectrum, some of the best combinations include the landscapes and sounds which are evaluated low at the single stimulus tests and also some worst combinations include the stream murmuring sound or wave sound which are highly ranked at the sound only tests, (4) since supersonic appears in many worst combinations, it is likely to bring about unpleasant effects on human psychology.
    Download PDF (8618K)
  • Takehito Hayami, Osamu Watanabe, Osamu Hiwaki
    2004 Volume 124 Issue 9 Pages 1712-1718
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    It is a useful method to estimate the position and the current distribution of peripheral nerves with magnetic field measurement as a non-invasive inspection. A simulation study was executed to estimate the electric current sources in a nerve fiber by the magnetic fields with L1-norm method. L1-norm method using magnetic fields usually estimated scattered dipoles over the space. In case of measurement of magnetic field around a limb, approximate direction of nerve fibers is known. Therefore this study proposed a method to estimate the current dipoles in a nerve fiber in an arm or leg by selecting the direction of current dipoles resulted from L1-norm method.
    Download PDF (602K)
  • Yuichiro Hashimoto, Masashi Kawasumi, Masao Saito
    2004 Volume 124 Issue 9 Pages 1719-1724
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    The effect of magnetic field on cell has long been investigated, but there are few quantitative investigations of the migration of cells. Cell-migration is important as one of the fundamental activities of the cell. This study proposes a method to evaluate quantitatively the cell-diffusion constant and the effect of static magnetic field on cell migration. The cell-lines are neuroblastoma (NG108-15), fibroblastoma (NIH/3T3) and osteoblastoma (MC3T3-E1). The static magnetic field of 30 mT or 120 mT is impressed by a permanent magnet in vertical or horizontal direction to the dish. It is shown that the cell-diffusion constant can represent the cell migration as the cell activity. It is found that the cell migration is enhanced by exposure to the magnetic field, depending on the kind of cell. It is conjectured that the effect of static magnetic field affects the cell migration, which is at the downstream of the information transmission.
    Download PDF (6184K)
  • Hideyuki Momoki, Hidenobu Arimoto, Naoto Kakuta, Yukio Yamada
    2004 Volume 124 Issue 9 Pages 1725-1730
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    We have proposed a new method of measuring temperature of media containing water by near infrared spectroscopy and multivariate analysis. The peak wavelength of the water absorption band at about 1450 nm shifts with the change in temperature. By measuring the wavelength shift of the absorption band we can measure the temperature of media containing water whenever it is transparent like water solutions or turbid like biological tissues. Fundamental experiments have validated the proposed method. The measurement accuracy has been found to be less than 0.1 °C for pure water, 0.3 °C for turbid media with transmission measurement, and 0.6 °C for turbid media with reflection measurement.
    Download PDF (5923K)
  • Motoshi Tanaka, Takayuki Miyashita, Hiroshi Inoue, Yoshitsugu Niiyama
    2004 Volume 124 Issue 9 Pages 1731-1737
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    As a fundamental study on the objective evaluation of TV picture degradation by electromagnetic noise with visual physiological information, the electroencephalogram (EEG) activity was measured when a still TV picture was degraded by the burst noise whose rms and duration were changed. A degradation was subjectively evaluated by three-grade impairment scale of “Not Annoying”, “Slightly Annoying”, and “Annoying”. Measured EEGs were analyzed by an averaging technique. In the results, the amplitude of the event related potential P300 becomes larger when the subjects feel the noise “Annoying”.
    Download PDF (372K)
  • Saeko Samejima, Masakazu Kobayashi
    2004 Volume 124 Issue 9 Pages 1738-1743
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    Autofluorescence (AF) measurement of human bronchus tissues are practically used for the detection of early stage lung cancers. Problems associated with the interference between the excitation light and the AF signal and the lack of the precise analysis of the AF characteristics limit the capability of the technique. In this study, we used a ultra-violet laser diode (UV-LD) and a ultra-violet light emitting diode (UV-LED) as the alternative light source for the AF measurement. An AF analysis system that can analyze the spectroscopic data of about 5mm×5mm with a spatial resolution of about 10 micro meters were established. The AF spectrum was measured for 512×512 pixels and the intensity mapping as a function of a certain emission wavelength was obtained. The numerical calculation of the data (including differential imaging and the chromaticity diagram projection) was performed and the precise feature of the AF was revealed.
    Download PDF (4716K)
  • Hirooki Aoki, Hidetoshi Nakamura, Masato Nakajima
    2004 Volume 124 Issue 9 Pages 1744-1752
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    Every conventional respiration monitoring technique requires at least one sensor to be attached to the body of the subject during measurement, thereby imposing a sense of restraint that results in aversion against measurements that would last over consecutive days. To solve this problem, we developed a respiration monitoring system for sleepers, and it uses a fiber-grating vision sensor, which is a type of active image sensor to achieve non-contact respiration monitoring. In this paper, we verified the effectiveness of the system, and proposed screening method of the sleep disordered breathing. It was shown that our system could equivalently measure the respiration with thermistor and accelerograph. And, the respiratory condition of sleepers can be grasped by our screening method in one look, and it seems to be useful for the support of the screening of sleep disordered breathing.
    Download PDF (8180K)
  • Osamu Sakata, Tsuyoshi Shiina, Yoichi Saito
    2004 Volume 124 Issue 9 Pages 1753-1758
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    In this paper, multidimensional directed phase analysis is proposed as a means of causality analysis for multiple time series, is proposed. The multidimensional directed phase is a phase with causality, which is associated with multidimensional directed coherence. Using multidimensional directed phase analysis, direction and time of a signal flow among multidimensional time series can be estimated. In two simulations, artificial time series have been analyzed by the new method to confirm its characteristics. In first simulation, accuracy of estimation of signal flow time has been investigated. Second, a complex signal flow pattern has been analyzed. Next, the multidimensional directed phase analysis and the multidimensional directed coherence analysis have been applied to EEG data of normal volunteer. As a result, we have got information differ from common learning which is the difference between the phase of a frontal alpha waves and an occipital is π.
    Download PDF (5060K)
  • Keiichi Fujita, Kazuto Tamura, Wataru Kaneko, Hiroaki Ishizawa, Eiji T ...
    2004 Volume 124 Issue 9 Pages 1759-1765
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    Recently, diabetics have been steadily increasing, because change of diet, lack of exercise, increase an alcoholic intake, and increase a stress. It is a very serious problem for us. About 23.6 millions of people in Japan approach the danger of diabetes. Therefore, it is necessary to get insulin injection. And they have to measure blood glucose again and again a day. So, they are burden too heavy.
    This paper describes a new noninvasive measurement of blood glucose based on optical sensing. This uses Fourier transform infrared spectroscopy of attenuated total reflection. Non-invasive measurement was carried out by using 3 methods. And standard error of prediction is about ±20mg/dl by 3 method. This paper also describes practical application of this method.
    Download PDF (5457K)
  • Takashi Seki, Kazuhiko Hamamoto
    2004 Volume 124 Issue 9 Pages 1766-1772
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    Recently, medical information system in dental field goes into digital system. In the system, X-ray image can be taken in digital modality and input to the system directly. Consequently, it is easy to combine the image data with alpha-numerical data which are stored in the conventional medical information system. It is useful to manipulate alpha-numerical data and image data simultaneously.
    The purpose of this research is to develop a new coding method for dental X-ray image. The method enables to reduce a disk space to store the images and transmit the images through Internet or LAN lightly. I attempt to apply multi-resolution analysis (wavelet transform) to accomplish the purpose. Proposed method achieves low bit-rate compared with conventional method.
    Download PDF (506K)
  • Kazuhiko Hamamoto, Motoyoshi Sato
    2004 Volume 124 Issue 9 Pages 1773-1779
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    3D imaging technique is very important and indispensable in diagnosis. The main stream of the technique is one in which 3D image is reconstructed from a set of slice images, such as X-ray CT and MRI. However, these systems require large space and high costs. On the other hand, a low cost and small size 3D imaging system is needed in clinical veterinary medicine, for example, in the case of diagnosis in X-ray car or pasture area. We propose a novel 3D imaging technique using 2-D X-ray radiographic images. This system can be realized by cheaper system than X-ray CT and enables to get 3D image in X-ray car or portable X-ray equipment. In this paper, a 3D visualization technique from 2-D radiographic images is proposed and several reconstructions are shown. These reconstructions are evaluated by veterinarians.
    Download PDF (6630K)
<Sound and Image Processing and Recognition>
  • Keiji Kurokawa, Shin Miura, Makoto Nishida, Yoichi Kageyama, Ikuro Nam ...
    2004 Volume 124 Issue 9 Pages 1780-1789
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    It is well known that temporal lobe in MR brain image is in use for estimating the grade of Alzheimer-type dementia. It is difficult to use only region of temporal lobe for estimating the grade of Alzheimer-type dementia. From the standpoint for supporting the medical specialists, this paper proposes a data processing approach on the automatic extraction of the intracranial region from the MR brain image. The method is able to eliminate the cranium region with the laplacian histogram method and the brainstem with the feature points which are related to the observations given by a medical specialist.
    In order to examine the usefulness of the proposed approach, the percentage of the temporal lobe in the intracranial region was calculated. As a result, the percentage of temporal lobe in the intracranial region on the process of the grade was in agreement with the visual sense standards of temporal lobe atrophy given by the medical specialist. It became clear that intracranial region extracted by the proposed method was good for estimating the grade of Alzheimer-type dementia.
    Download PDF (8852K)
<Control and Measurement>
  • -A concurrent Quantity Processing Mechanism in Actual Computation-
    Li-qun WANG, Masao Saito
    2004 Volume 124 Issue 9 Pages 1790-1794
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    We used 1.5T functional magnetic resonance imaging (fMRI) to explore that which brain areas contribute uniquely to numeric computation. The BOLD effect activation pattern of metal arithmetic task (successive subtraction: actual calculation task) was compared with multiplication tables repetition task (rote verbal arithmetic memory task) response. The activation found in right parietal lobule during metal arithmetic task suggested that quantitative cognition or numeric computation may need the assistance of sensuous convert, such as spatial imagination and spatial sensuous convert. In addition, this mechanism may be an ’analog algorithm’ in the simple mental arithmetic processing.
    Download PDF (4605K)
Special Issue Letter
<Biomedical Engineering>
Paper
<Integrated Electronic-Circuits>
  • Mitsuya Otake, Kazuyuki Wada, Yoshiaki Tadokoro
    2004 Volume 124 Issue 9 Pages 1797-1804
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    Design of integrated wave filters with effect due to parasitics relaxed is proposed. Improvement of frequency characteristics is achieved by using current-mode circuits and appropriately transforming the transfer function of a wave unit. Circuitry of two building blocks one of which follows another in a wave filter is also simplified for eliminating parasitics. Moreover small-signal analysis leads to a design guidance that the gate-source bias voltage of a MOSFET which operates as an equivalent resistor should be large. Two wave filters are designed based on the guidance as examples and validity of the guidance is verified with SPICE.
    Download PDF (429K)
<Multimedia Technology>
  • Yasuhiro Suda, Mitsunobu Kamata, Akihiko Sugiura
    2004 Volume 124 Issue 9 Pages 1805-1811
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    Presently, by using motion capture system for generating human body movement in CG, it is possible to take in movement correctly. However, motion capture system requires expensive equipment and cannot be used easily. Therefore, we developed a system which can compound and make a human body movement animation simply. In addition, a human body movement animation needs human body model. Action Unit is used for this model. Action Unit takes human’s muscles into consideration, and it has arranged muscles so that it may become the right position anatomically. We evaluated an usability of the depeloped system, and checked its validity.
    Download PDF (259K)
<Sound and Image Processing and Recognition>
  • Yohei Hamada, Shingo Ando, Masato Nakajima
    2004 Volume 124 Issue 9 Pages 1812-1819
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    When we take a picture of a scene including a bright area and a dark area with a typical off the shelf camera, we generally get pictures with very dark shadows and bleached out areas. These effects are caused by the low quality of the camera. Ordinary cameras have a narrow dynamic range, so they can’t take quality pictures under poor conditions like into the light. We are developing a camera that can take quality pictures without dark shadows nor bleached out areas under any situations. We change the structure of imaging equipment. Ordinary cameras have one CCD. We increase the number of CCDs to two. Therefore, one is for taking an underexposed picture and the other is for taking an overexposed picture. The underexposed picture provides a high-quality image of the bright part. The overexposed picture provides a high-quality image of the dark part. Firstly, these two pictures are taken at the same time. Secondly, the two pictures are combined taking the high quality areas from both at the most appropriate rate. As a result, we can get a compounded picture without dark shadows nor bleached out areas. In this paper, we prove this method could be effective under our algorithm.
    Download PDF (5508K)
  • Hiroya Igarashi, Hideo Saito, Shinji Ozawa, Keiichi Yamamoto, Hiromits ...
    2004 Volume 124 Issue 9 Pages 1820-1826
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    In this paper, we propose a system for switching operation of vehicle instrument panel by hand gesture. When driving a vehicle, a driver often needs to watch instrument panels for switching operation, such as an air-conditioner and an audio system, although such operation might reduce visual attention of vehicle-front view for safety driving. The proposed system makes a driver to perform such switching operation by hand gesture without watching the instrument panel. This system captures images of a driver’s hand with the camera attached on the ceiling inside of a vehicle. The angle of the index finger are detected by processing the captured image sequence. The detected hand’s motion is recognized as switching operations instead of the switching operation on the instrument panel. This system has two modes for controlling audio and air-conditioner. The first mode is for selecting the control function, such as volume of audio, temperature of air-conditioner, and strength of the airflow. Such selection is performed by pushing gesture by the index finger. Next, the value of the selected function is adjusted by the position of the hand. Using this system, the driver can control the interior systems without watching the instrument panel, so that the visual burden at the time of the switch operation can be reduced.
    Download PDF (280K)
<Control and Measurement>
  • Isao Yamada, Satoshi Kumagai, Toyohiko Hirota
    2004 Volume 124 Issue 9 Pages 1827-1833
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    In order to comply with rapid changes of business environment, companies or organizations continuously modify contexts of their business process. Different data model will be built to support business process to be changed. In this paper, we introduce PCM (Property Centric Modeling) which facilitates data model design in accordance with business process. Context is often appeared in business process as function, persons, facilities, or materials. PCM associates each context with set of properties to build data model from those properties. Resulting data model is flexible enough for the change of context. Specific procedure of the method is illustrated, taking a sensorchip bonding process as example. Comparisons with other methodologies are also presented.
    Download PDF (161K)
  • Osamu Kubo, Yasuhiro Kobayashi
    2004 Volume 124 Issue 9 Pages 1834-1840
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    Weather derivatives are getting to be powerful tools for weather risk hedging. A popular method which draws out valid prices of weather derivatives is a stochastic modeling approach. In the method, expected payoffs of weather derivatives based on stochastic weather models are regarded as their valid prices. Although useful stochastic models of temperature have been shown, stochastic models of daily rainfall are still being developed. Therefore, it is considered that pricing of daily rainfall derivatives is difficult. This paper shows a new stochastic daily rainfall model for pricing daily rainfall options. The new model in which a modified geometric distribution model is applied can express stochastic features of daily rainfall. Furthermore, this paper also shows that the combination model of the Markov chain rainy day model and the new model can express stochastic features and risks of daily rainfall option payoffs.
    Download PDF (5678K)
  • Hitoshi Takata, Toshinori Nawata
    2004 Volume 124 Issue 9 Pages 1841-1846
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    In this paper a nonlinear feedback control called augmented automatic choosing control (AACC) for a class of nonlinear systems with constrained input is presented. When designed the control, a constant term which arises from linearization of a given nonlinear system is treated as a coefficient of a stable zero dynamics. Parameters of the control are suboptimally selected by minimizing the Hamiltonian with the aid of a genetic algorithm.
    Download PDF (165K)
  • Yoshida Mitsuo, Takehiro Mori, Yasuaki Kuroe
    2004 Volume 124 Issue 9 Pages 1847-1852
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    In this paper, we propose a condition on global asymptotic stability for recurrent type complex-valued neural networks with a class of activation functions. Those networks have states, connection weights, and activation functions, which are all complex-valued. Such networks have been studied concerning their abilities of information processing, because of their attractive features which are not existent in real-valued counterparts. In particular, the activation function of a complex-valued neuron is an important factor that characterizes the dynamics of complex-valued neural networks. We propose a class of activation functions which ensure the global asymptotic stability, together with a condition on the connection weights. Furthermore, we apply the proposed dynamical networks and the stability condition to a convex programming problem with nonlinear constraints.
    Download PDF (162K)
<Neural Network, Fuzzy and Chaos Systems>
  • Masaharu Tanaka, Masanobu Mizoguchi, Isao Takami
    2004 Volume 124 Issue 9 Pages 1853-1860
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    When Genetic Algorithm (GA) is applied to actual engineering problems, for example applied to optimization of design parameters, the searching time is usually huge because the fitness is calculated by repetitive simulation or analysis, which needs a large amount of calculation time. In order to shorten the searching time, a method called fitness prediction GA has been proposed. It reduces the calculation time by predicting fitness instead of actually calculating it, and eventually shortens the searching time. In this paper, we propose a new method, Dual Population GA (DPGA), which employs both real and virtual populations. Real populations have actual fitness value, and virtual ones have predicted one. DPGA can prevent the decline of performance caused by prediction errors, which may happen in fitness prediction GA, by appropriately migrating virtual populations into real ones and accelerating evolution. A fitness predictor based on a neural network is also proposed in this paper. Through computer simulations, DPGA is proved to be able to improve the searching performance of fitness prediction GA.
    Download PDF (188K)
  • Kenichi Ida, Ryouhei Suga, Mitsuo Gen
    2004 Volume 124 Issue 9 Pages 1861-1867
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    The greedy algorithm is one of solution methods for knapsack problem. Although this algorithm does not necessarily obtain the optimal solution, it can obtain a good solution in short time. We consider this algorithm is very effective at judging the importance of each item.
    In this paper, we propose a new genetic algorithm for solving a knapsack problem. The algorithm can adjust a search area in consideration of the stability of each item which can obtain from the greedy algorithm. Moreover, we apply the proposed method to a multi-objective problem and a large-scale problem, and test the effectiveness.
    Download PDF (333K)
<System Engineering>
  • Shigeki Ishitani, Sadamichi Mitsumori
    2004 Volume 124 Issue 9 Pages 1868-1873
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    This paper analyzes a relation between level of total production inventory and effective production capacity in a production process, and proposes a method of controlling effective production capacity in a business fluctuations period. They are based on the property: “effective production capacity = capacity of production facilities - total changeover loss", where changeover loss is a decreasing function of level of total production inventory.
    Although the effective production capacity can be controlled by an increase or decrease of production facilities, its cost is too expensive. It must be done only in the case that an increase or decrease in demand is confirmed. Therefore, the effective production capacity must be controlled by the level of total inventory.
    In a business recovery or recession period, two kinds of production time have to be controlled. The first the production-time control for the increase or decrease in demand, and the second is that for an increase or decrease in the level of total inventory. Even if an increasing or decreasing rate in demand is constant, the value of the total control time decreases at first and then increases exponentially. This is the reason why inventory control is difficult in a business fluctuations period.
    Download PDF (6175K)
  • Hiroshi Murata, Takashi Onoda, Katsuhisa Yoshimoto, Yukio Nakano, Syuh ...
    2004 Volume 124 Issue 9 Pages 1874-1880
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    This paper presents applying results of four estimation algorithms of non-intrusive monitoring system for real household. We conclude that all algorithms have practicable ability. 1) support vector machine(SVM): SVM was used to estimate ON/OFF states for fluorescent and refrigerator. SVM has the performance equivalent to best performance of sigmoid function networks(SFN). However, SVM has high estimating ability constantly. 2) RBF networks(RBFN): RBFN was used to estimate power consumption for air conditioner. RBFN has the performance equivalent to best performance of SFN. However, RBFN has high estimating ability constantly. 3) step change detection method(SCD): SCD was used to estimate ON/OFF states and power consumption for IH cooking range. SCD does not need the necessary learning process for SFN and has higher estimating ability than SFN. 4) spectrum reference method(SRM): SRM was used to estimate working conditions for rice cocker and washing machine. SRM is able to estimate these working conditions that cannot be estimated by earlier methods.
    Download PDF (330K)
  • Kei Sakuma, Eitaro Aiyoshi
    2004 Volume 124 Issue 9 Pages 1881-1887
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    This paper is concerned with a computational method to search for all local minima. In the proposed method, a coupled multi-agents dynamical model is used, in which a single agents moves on the gradient model under repulsion from other agents as well as under repellent from trapped agents on a local minimum. These repulsion and repellent mechanisms are effective to avoid from the discovered local minima and to search for the undiscovered local minimum. Lastly, the behavior of this algorithm is investigated with simulations for several examples.
    Download PDF (231K)
  • Takashi Okamoto, Eitaro Aiyoshi
    2004 Volume 124 Issue 9 Pages 1888-1895
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    Optimization methods by using chaos dynamics are interesting as a class of global optimization methods by which the global minimum can be obtained without trapping in local minima. The chaos dynamics are classified into discretized gradient models and continuous dissipative models with a nonlinear damping term. In this paper, optimization problems with hyperspherical constraints are considered in order to present nonlinear dissipative dynamics embedded in their constraints. As the nonlinear dissipative dynamics, Fujita-Yasuda’s Model(2) and Tani’s Model(3) are adopted. Especially, their revised models are proposed newly for the hyperspherical constraints. The numerical simulations for a few constrained optimization problems demonstrate effectiveness of presented constrained global optimization methods.
    Download PDF (615K)
<Software and Information Processing>
  • Mitsukuni Matayoshi, Morikazu Nakamura, Hayao Miyagi
    2004 Volume 124 Issue 9 Pages 1896-1906
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    We propose a new 2-opt base method as a local search approach used with Genetic Algorithms (GAs) in Memetic Algorithm. We got a hint from the fast 2-opt method and devised the new 2-opt method. The main different point is such that our method exchanges genes by using histories of contributions to fitness value improvement. The contribution level is represented by the value `Priority’.
    In computer experiment, Quadratic Assignment Problem (QAP) instances are solved by GA with the 2-opt method(First Admissible Move Strategy, the Best Admissible Move Strategy), the fast 2-opt, and our proposed method for comparative evaluation. The results showed that our improved method obtained better solutions at ealier generation of the GA and our method required less computation time than the others at some upper bound value of appropriate `Priority’ setting values. Specially, at the average elapsed time of the fast 2-opt method’s 1000th generation, the exact solution findings of ours is more than the others.
    In further experiment, we observe that the searching capability depends on the number of levels of `Priority’. The ratio between two different Priority level sets becomes 1.59 in computation time in solving problem instance “char25a". This characteristic is shown to be statistically significant in ten instances among eleven.
    Download PDF (513K)
  • Yoshikazu Ishii, Tarou Takagi, Kouji Nakai
    2004 Volume 124 Issue 9 Pages 1907-1913
    Published: 2004
    Released on J-STAGE: December 01, 2004
    JOURNAL FREE ACCESS
    In order to analyze the output of the intrusion detection system and the firewall, we evaluated the applicability of ICA(independent component analysis). We developed a simulator for evaluation of intrusion analysis method. The simulator consists of the network model of an information system, the service model and the vulnerability model of each server, and the action model performed on client and intruder. We applied the ICA for analyzing the audit trail of simulated information system. We report the evaluation result of the ICA on intrusion analysis. In the simulated case, ICA separated two attacks correctly, and related an attack and the abnormalities of the normal application produced under the influence of the attach.
    Download PDF (7192K)
Letter
<Biomedical Engineering>
<Control and Measurement>
<Software and Information Processing>
feedback
Top