電気学会論文誌C(電子・情報・システム部門誌)
Online ISSN : 1348-8155
Print ISSN : 0385-4221
ISSN-L : 0385-4221
119 巻, 6 号
選択された号の論文の16件中1~16を表示しています
  • 成田 憲一, 本間 規泰, 後藤 幸弘, 竹内 伸直, 中鉢 憲賢
    1999 年 119 巻 6 号 p. 655-661
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    The meteorological elements and the atmospheric electric field have been observed continuously for monitoring thunderclouds at two observation sites. In this paper the variations of these elements are described and analyzied qualitatively using auto-and mutual-correlation method. From the obtained results when the air current begins to flow downward the wind speed increases and the atmospheric temperature rises simultaneously. After about ten minutes the decending air current starts and the atmospheric pressure increases. This result agrees with the growing and vanishing model of a thundercloud.
  • 高原 健爾, 若松 秀俊, 宮里 逸郎
    1999 年 119 巻 6 号 p. 662-667
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    This study is to design a system to control oxygen saturation of arterial blood automatically taking into consideration the clinical care of hypoxemia patients. For this purpose, the control system is based on the adaptive pole-placement method, choosing oxygen concentration of inspired air and oxygen saturation of arterial blood as the manipulating and controlled variables, respectively. The controlled respiratory system is assumed to be described by a first-order linear model. The control system is synthesized taking into account the practical artificial respiration using a computer-controlled respirator, a pulse oximeter in clinics. For the representation of the respiratory system, the individuality and the change in its characteristics of subjects are regarded as the changing parameters of the mathematical model in the concept of the adaptive control method. The proposed system is shown stable enough and clinically safe by the simulation experiments, which yields its effectiveness both at conditions of rest and some state change affected by metabolic rate change corresponding to oxygen consumption rate change.
  • 面白い・楽しいの場合
    堀 政重, 小関 修, 横山 清子, 渡辺 與作
    1999 年 119 巻 6 号 p. 668-675
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    The purpose of this study is to propose the estimating method of the intensity of human emotion for fun and pleasure using facial expression and physiological data. The index of facial expression is the transverse width of mouth and the indices of physiological data are heart rate and skin conductance level. The three indices were measured while subjects were seeing comedy movies. The subjects reported their evaluation for fun and pleasure with four intensity levels (0, 1, 2 and 3) after movie. Using the three indices, the estimation formula based on multiple regression model was derived. The estimated intensity was compared with reported one. The mean error of estimation was about 0, 3. It was concluded that the proposed method would be able to estimate the intensity of human emotion for fun and pleasure.
  • 小林 真也, 左木 茂克
    1999 年 119 巻 6 号 p. 676-683
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    Task Scheduling is key for high speed processing with a multiprocessor system. Scheduler should take consider in not only task size but also communication overhead. Effect of communication overhead is not constant. If a task is smaller than others, communication over head has more effective to this task than to others. By the way, traditional scheduling algorithm uses idle processors aggressively when idle processors exist. It increases communication between processors, and then finally processing time with multiprocessor overcomes processing time with single processor.
    In this paper, we propose new task scheduling algorithm. This algorithm is improved version of CP/RCO. This can consider relative relation between task size and communication overhead, and also pending allocation if other busy processor to execute task faster exists. We present that our proposed algorithm is more efficient than traditional algorihtms, and that computaion complexity of our algorithm is O (mn2).
  • 結城 健太, 難波 雪, 宮崎 道雄
    1999 年 119 巻 6 号 p. 684-690
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    Recently, the research of chaos has come into vogue. Chaotic behavior is observed in many different fields and applications of chaos for these situations are required.
    In this paper, the 3-dimentional chaotic trajectory is controlled by using poincare surface and the plural Unstable Periodic Regions (UPR) when the chaotic map is unknown. Furthermore we propose a new concept of Double Poincar_??_ map method for the estimating chaos trajectory and can get the good estimated value. Finally, we confirm the effectiveness of this method with numerical experiments.
  • 入室管理システムへの応用
    陳 文杰, 岡本 教佳, 南 敏
    1999 年 119 巻 6 号 p. 691-699
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    This paper proposes a new algorithm of the automatic personal identification and the practical application of the algorithm to the entrance control system using extracted contour lines and feature points from human face profiles and summarizes the simulation results of the identification. As a decision function for the identification, we use a weighted distance between the registered 19 dimensional feature vector and the 19 dimensional feature vector of the input profile. The components of the 19 dimensional feature vector are the distances between two feature points and the angles between two lines connecting three consecutive feature points. The 11 feature points are extracted from a contour line of the input profile expressed by Freeman's chain code using digital curvatures of the line. The effects of deformation of profiles caused by face panning and tilting and mouth opening upon identification accuracy have been investigated in details. To overcome the deformation effects we propose to register three profiles per person: a normal head position profile, a tilted profile and a panned profile. The simulation results by 74 subjects show that the identification accuracy for same persons is 80.6_??_97.2% and discrimination accuracy for different persons is 98.1_??_91.1%. This proves the superiority of the proposed algorithm.
  • 富永 昌治, 鹿島 恵
    1999 年 119 巻 6 号 p. 700-707
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    A lighting system is proposed for measuring color images under a variety of illumination. This system is constructed with halogen lamps, color filters, white diffusion filters, dimmers, and a personal computer as a controller. Colored light with continuous spectral power distribution is produced based on the additive color mixture of RGB primary lights. First, we describe the method for producing light with a target color stimulus. The basic procedure is performed in two steps of (1) a XYZ-RGB color coordinate conversion and (2) correction of nonlinearity, A practical procedure is presented for easily producing the colored light with any chromaticity coordinates (x, y) specified within the color gamut. We show a method of improving low luminance of the blue region corresponding to high color temperature. Next, the performance of the proposed lighting system is investigated. The accuracy of the colored light is examined on a color difference, and the performance is compared with that of an incandescent lighting system. The goodness is confirmed on the accuracy and the color gamut. Finally, we describe an application of the lighting system to computer vision. The illuminant-spectral distribution and surface-spectral reflectances are estimated from the image data photographed by a multi-band camera.
  • 直江 秀憲
    1999 年 119 巻 6 号 p. 708-713
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    The present paper proposes a modeling method which describes and predicts time series phenomena in a qualitative manner. For the qualitative description, time series data are characterized by the first and second order coefficients which are the outputs of a time dependent polynomial filter. In order to predict by the neural net, the coefficients are transformed and normalized to [0, 1] space. From the series of these normalized coefficoients, a few teaching patterns for neural net are extracted by clustering algorothm. And by error back propagation algorithm, the future state of the time series phenomena is obtained as symbols. Because the stock price does not necessarily walk randomly, the stock prices are tried to identify and predict as the time series data. Comparing the proposal method with convential autoregressive model, it is clarified that the former has even high precision concerning with the predicted pattern by the correlation coefficients.
  • Akio Koyama, Leonard Barolli, Said Mirza, Shoichi Yokoyama
    1999 年 119 巻 6 号 p. 714-723
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    The Asynchronous Transfer Mode (ATM) technique has been accepted as a basis for the future B-ISDN networks. With the B-ISDN/ATM goals of supporting diverse services and traffic mixes, and of efficient network resource engineering, the design of the traffic control becomes a difficult task. The ATM Forum has chosen the rate-based scheme as an approach to congestion control for the Available Bit Rate (ABR) services. So far, many rate-based congestion control schemes have been proposed. Some of them use the negative feedback rate control. Therefore, if all notification cells in backward direction experience extreme congestion, the overall network congestion collapse may happen. Others use the positive feedback rate control, but the unfair distribution of available bandwidth among Virtual Connections (VCs) may occur. To resolve the problems of the existing rate-based control scheme, in this paper, we propose a new congestion control scheme called Exact Explicit Rate Indication (EERI) scheme. In the EERI scheme, the network explicitly informs each source the exact rate they should transmit a message. The network determines the rate of each source based on the fair rate setting procedure. The rate determination mechanism doesn't require per-connection queuing or per-accounting in the network, which are considered expensive techniques with the current hardware technology. By using the Benchmark Fairness Configuration (BFC), we show that the proposed scheme can achieve an excellent fairness performance among connections existing in the network. Furthermore, in the EERI scheme there are not oscillations in the rate allocation. In the transient behavior, the EERI scheme offers fast access to available bandwidth, which is a sensible requirement particularly in the LAN and MAN environments.
  • 中野 和司, 田原 鉄也, 新 誠一, 豊田 幸裕
    1999 年 119 巻 6 号 p. 724-733
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    The wavelet analysis is superior to the Fourier analysis in the detectability of temporal (or spatial) information of signals. Therefore, the use of the analyzing wavelet which exists locally in the frequency domain makes it possible to estimate the time delay and the frequency response of the process transfer function in a specific time domain. First, we present a time-delay identification method by using the wavelet transform of the cross-correlation of input/output (I/O) data. Secondly, we present a method for obtaining the frequency response of the process to be considered as a time-varying system by directly using the wavelet transform of the auto-/cross-correlations of the I/O data. Thirdly, we determine the parameters in the transfer function by using the curve-fitting technique. Finally, we demonstrate the feasibility and validity of our identification procedure through some simulations and experiments for an actual boiler process.
  • 和田 信敬, 佐伯 正美
    1999 年 119 巻 6 号 p. 734-739
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    In the practical system, there exist saturation limitations on the control inputs. Difference between the output of a controller and the control input causes a windup phenomenon and gives adverse effects on the control performance. In this paper, we study a design method of a compensator which improves anti-windup property and robust stability. Design of anti-windup compensator is converted into an optimization problem with constraints described by two matrix inequalities that represent anti-windup and robust stability conditions, respectively.
  • 直江 秀憲
    1999 年 119 巻 6 号 p. 740-746
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    The present paper proposes a portfolio selection method on the concept of neural network, which selects a few stocks with low risk and high returns among the many. It was difficult to solve this problem even with a large computer, because the number of the feasible combination of the stocks is enormous. Applying the concept of neural network, quasi-minimum points of the energy function are searched and the quasi-optimal stocks are obtained. The measure of the performance is the deviation from the mean. Then, the performance of the solution by the simulated annealing method using the Hopfield network is compared with that by the simulated annealing method with the Lagrangian multiplier, by the greedy method, by the all searching method and by the random method. It is concluded that the performance of these several methods is almost the same. The risk and the return of the portfolios have the trade-off relation. Consequently, it is possible to apply the simulated annealing method using the Hopfield network in order to select portfolios
  • 作田 幸憲, 古川 誠, 関根 好文
    1999 年 119 巻 6 号 p. 747-752
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    Measuring the phase noise using a PLL is known as a highly precise method. But the method can't be used when the stability of a test oscillator is higher than the stability of a reference frequency source in the PLL.
    So we propose a new method that reduces the phase noise in the reference frequency source. This method adds outputs of m PLLs locked by the standard oscillator which is the reference frequency source. In this paper, we clarify theoretically that L(f) can be improved by 10 logm dB in Fourier frequency range which is higher than the PLL's natural frequency and substantiate this experimentally.
  • 山下 育男, 猪口 勝司, 青海 恵之
    1999 年 119 巻 6 号 p. 753-759
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    Application of semiconductor optical amplifiers to a 2.5Gbit/s×4 WDM repeaterless transmission system using the 1.3-μm wavelength band has been studied. Measured gain characteristics and bit error rate properties for 2.5Gbit/s signals amplified by the SOA's are found to be enough for post and pre-amplifier of the 4-channel WDM. The repeaterless transmission experiment over a 100-km SMF shows the allowable line loss of 32.8dB, and the same results are obtained by a experiment using fibers constructed in the field. Problems of applying SOA's to the WDM system such as its stability degradation due to polarization fluctuation and the influence of interference among channels are studied. When suitable margin is considered in the system, the influence of a polarization fluctuation is negligible. However, in order to avoid interference among channels, input optical power of the SOA should be around -22dBm/ch for 4-channel WDM.
  • 岩本 元
    1999 年 119 巻 6 号 p. 760-761
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
  • Motoshi TANAKA, Hiroshi INOUE
    1999 年 119 巻 6 号 p. 762-763
    発行日: 1999/06/01
    公開日: 2008/12/19
    ジャーナル フリー
    The human footsteps recognition (walk-recognition) is discussed. The pitch frequency related with walking pace, and the first peak frequency of a footstep, are characterized by frequency analysis, and used as a feature vector for the recognition. A simple recognition method comparing with the Euclidean distance are used. The recognition rate is approximately 83%, and then the feasibility of walk-recognition is confirmed.
feedback
Top