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
124 巻, 4 号
選択された号の論文の17件中1~17を表示しています
論文
<医用電子・生体工学>
  • 久野 弘明, 高島 充, 大川井 宏明
    2004 年 124 巻 4 号 p. 935-940
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    In this study, the possibility of the measurement of respiration, heart beat, and body movement on a bed was examined using the dynamic air-pressure sensor aiming at a daily health monitoring. The dynamic air-pressure sensor measures vital information using a change of air pressure. Twelve healthy volunteers participated in this study. The dynamic air-pressure sensor was installed under the bed mat and respiration and heart beat information were measured. This information was compared with the standard waveforms obtained from respiratory belt transducer and the electrocardiograph. As a result, both waveforms demonstrate a high correlation, and confirmed the validity of this method. A change of waveform and a quantitative evaluation of respiration, heart beat, and body movement measured from during sleep using this sensor can be useful for a daily health monitoring.
<通信・ネットワーク>
  • 渡邉 勝秀, 林 海, 山下 勝己
    2004 年 124 巻 4 号 p. 941-946
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    This paper describes a new solution to the channel intersymbol interference and the nonlinear distortion by means of adaptive array antenna and radial basis function (RBF) equalizer. RBF equalizer can reduce the influence of intersymbol interference and additive noise in the digital communication system. It is also well known to eliminate the nonlinear distortion effectively. But the conventional RBF equalizers including our previous proposed equalizer, spatial and temporal RBF equalizer using adaptive array antenna, need training data to learn channel states, which incurs decline in the data rate. In this paper, without training data, we construct a blind spatial and temporal RBF equalizer using genetic algorithm. Furthermore, we also describe a method of synthesizing Bayesian decision variables of RBF equalizers in order to obtain better performance. The RBF equalizer based on the proposed method can present a remarkable improvement in BER characteristics and its performance is demonstrated by computer simulations.
<音声画像処理・認識>
  • 後藤 富朗, 奥村 剛史, 北村 正, 早原 悦朗
    2004 年 124 巻 4 号 p. 947-952
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    RGB is the standard color format for most digital video input and output devices, such as display monitors. However, each component consisting of a natural image has a lot of redundancy. So, a color conversion, which is RGB to YCrCb, has been being used to reduce its redundancy. However, each component of YCrCb signals has no correlation under the higher compression rate, so compression performance is decreased. So, we propose a new color conversion. By implementing the proposed method, which exploits the correlations between the color components, the PSNR of traditional way is higher than that of proposed way under low compression performance. However, it can not be found that the difference of reconstructed images quality is invisible. And the PSNR of traditional way becomes worse quickly under a high compression performance, however, that of proposed way becomes worse slowly, therefore proposed way can produce a better quality of reconstructed image in a low bit-rate. As a result, the examples demonstrated that a better performance of color image coding is obtained.
  • 溝尻 勳, 近藤 貴之
    2004 年 124 巻 4 号 p. 953-958
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    For characterization or discrimination of a 2D-man-face image, it is usually examined a face-outline and features, such as eyes, nose, lip, and etc.
      A standard face-outline is easily obtained by averaging the outline curve function defined in a properly normalized domain. But it is difficult to obtain a standard feature, it has to represent and average displacement of features, moreover, it may be needed to average of their time variations. We represent the correspondence of one image to the other displacement of features as a 2D-warping function using the face-outline curve coordinate system. This coordinate system is spanned by analogues of face-outline curves and arguments around the origin. In this method, the 2D-face image is automatically normalized by the face-curve, and mapped on to the disc. Adaptation of this coordinate system not only saves calculation but also makes easy to determine the warping function.
      We describe successive procedures and examine an ability of this method.
  • 岩片 智, 味岡 義明, 萩原 将文
    2004 年 124 巻 4 号 p. 959-965
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    In this paper, we propose an algorithm extracting character domains from scenery images. This algorithm works under a severe constraint: each pixel in a processed image must be derived from only information of the neighboring pixels. This constraint is very important for a low cost device such as a mobile camera, because it makes possible to process each pixel in parallel. The proposed algorithm deals with color information based on physiological knowledge, and uses size of regions for judging whether a region is characters or not. The proposed algorithm is represented by local and parallel image processing and has been tested using 100 scenery images. The result shows that the proposed algorithm can extract character regions in a rate of more than 90%.
<制御・計測>
  • 奥谷 巖, 竹内 正軌, 高瀬 達夫
    2004 年 124 巻 4 号 p. 966-973
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    Already proposed was a method using the Kalman filter which aimed to estimate the land-cover condition of a small area of the order of 200m×200m from remotely sensed data based on category decomposition principle. This method utilized only average values of band data for both categorical reflectance spectra and observation quantity over the small area. In the present paper we put forward a new land-cover estimation method using the extended Kalman filter which can take account of covariance features of categorical reflectance values as well as pixels brightness variation within the small area. The method is tested over the 4km×8km area drawn out from Odawara city for its validity verification. The results show that the new method outperforms the Kalman filtering method by 9.6% in terms of root mean square error in five land-cover category estimation and by 17.2% in thirteen category estimation.
  • 盛重 肇, 土井 繁, 田中 正吾
    2004 年 124 巻 4 号 p. 974-980
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    In diagnosing bearings, acceleration or eddy-current sensors is frequently used. However, sometimes there are objects to which the sensors cannot be applied because of their structures and the cost to be spent. As such an object, there is a sump pump used in a waste liquid transportation in nuclear power plants, where the shaft is put in the waste liquid.
      The paper proposes a signal propagation model based diagnosis method which monitors the anomaly of the bearing indirectly by measuring the shaft vibration. The vibration measurement is achieved by using an ultrasonic sensor attached on the casing surface with a magnet, so that the system serves as a portable unit which can reduce the monitoring cost remarkably. The effectiveness of the method is demonstrated with simulations and an experiment.
  • 池之上 正人, 金江 春植, 楊 子江, 和田 清
    2004 年 124 巻 4 号 p. 981-988
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    Many identification methods are based on the assumption that input measurement is noise-free. However, this condition is not satisfied in most practical situations. In the presence of input noise, those methods have been shown to give erroneous results. Bias-compensated least-squares method is a consistent estimation method for unknwon parameters of pulse transfer function in the presence of input and output noise. The method is based on compensation of asymptotic bias on the least-squares estimates by making use of noise variances estimates. In this paper, a new type of noise variances estimation method is proposed to overcome drawbacks of previously proposed methods. The main feature of proposed method is to introduce a generalized least-squares type estimator in order to estimate noise variances. The results of a simulated example indicate that the proposed method provides good estimates.
  • 久保 智裕
    2004 年 124 巻 4 号 p. 989-994
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    In the finite dimensional case, the linear quadratic regulator (LQR) is known to have good robustness properties. These properties are independent from choosing the weighting matrices in the cost functional. So, if a control system belongs to a class of LQRs, it is assured to have these robustness properties. In this paper, a method is proposed to construct a robust feedback controller for uncertain linear systems with time-delay of retarded type. The feature of this method is that the resulting closed loop system is assured to be an LQR for some cost functional. The cost functional contains some uncertainty, and it absorbs the uncertainty in the plant parameters so that a fixed feedback law is always optimal. The feedback gain is calculated with a solution of some linear matrix inequalities, and it is shown that the resulting closed loop system is an LQR for some cost functional, so it has the same robustness properties with the finite dimensional LQRs. Introducing an auxiliary system, it is possible to assign the degree of exponential stability of the closed loop system. A numerical example is given to demonstrate the effectiveness of this method.
<システム>
  • 中村 次男, 鈴川 敦之, 冬爪 成人, 笠原 宏, 田中 照夫
    2004 年 124 巻 4 号 p. 995-1003
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    As for the next generation of ultra large-scale integrated circuit called system LSI and system in package, the difficulty in designing many large circuit blocks has been pointed out. Therefore, it is being searched for new ways to cope with designing them.
      In this paper, we propose a method of hardware modularization, each of which has the function of judging necessary operation without any control from the outside, and without any limitation of the operational precision, i. e. an autonomous module. We also discuss how to standardize the module interface for ease-of-use.
  • 高森 真仁, 小圷 成一, 濱上 知樹, 平田 廣則
    2004 年 124 巻 4 号 p. 1004-1011
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    In this paper we propose a modified gene coding and an evolutionary construction in consideration of failure in evolutionary construction of Block-Based Neural Networks. In the modified gene coding, we arrange the genes of weights on a chromosome in consideration of the position relation of the genes of weight and structure. By the modified gene coding, the efficiency of search by crossover is increased. Thereby, it is thought that improvement of the convergence rate of construction and shortening of construction time can be performed. In the evolutionary construction in consideration of failure, the structure which is adapted for failure is built in the state where failure occured. Thereby, it is thought that BBNN can be reconstructed in a short time at the time of failure. To evaluate the proposed method, we apply it to pattern classification and autonomous mobile robot control problems. The computational experiments indicate that the proposed method can improve convergence rate of construction and shorten of construction and reconstruction time.
  • Hugang Han
    2004 年 124 巻 4 号 p. 1012-1020
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    Based on the Lyapunov synthesis approach and regarding the fuzzy systems as approximators to approximate the unknown functions in the system to be controlled, several adaptive fuzzy control schemes have been developed during the last decade. Actually, these schemes have been applied only to simple classes of nonlinear systems. In the concrete, (i) most of them just consider SISO systems (which can avoid the challenging of the coupling between control inputs); (ii) the upper bounds of uncertainties, and the reconstruction errors between the best approximators and their corresponding functions to be approximated are assumed to be known (in this way, the traditional adaptive methods or robust methods could be utilized straightforwardly). This paper develops a design methodology that expands the class of nonlinear systems to MIMO systems, the above restrictive assumptions can be relaxed by using an unique way to deal with the uncertainties and the reconstruction errors. The overall adaptive scheme is shown to guarantee the tracking error, between the outputs of system and the desired values, to be asymptotical in decay.
  • 小林 真也, 久原 俊介, 清家 悠, 樋上 喜信
    2004 年 124 巻 4 号 p. 1021-1028
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    We have implemented commends of the Autonomous Load Distribution Algorithm(ALD) for general users. ALD is one of inter-node protocol that distributes tasks among computers by negotiation between computers. We had developed and implemented ALD on multi-computer system in the past. But aim of this implementation was just for evaluation. Unfortunately, that was not easy to use for general users.
      In this paper, we have proposed five commands to use ALD for general users, and implemented these commands. Four of these are barn new commands, and one is same as command that was developed for evaluation in past. User can indicate task distribution in explicit way or in implicit way depending his/her demand. And user can similarly indicate task no-distribution in explicit way or in implicit way. In other words, user can obtain high performance without recognizing existence of ALD.
      Implementation is based on shell’s alias function and trap mechanism. Therefore, this system has higher portability. Alias function is simple, but powerful mechanism. But, it has problem of difference of name and body. This problem is that name indicates incorrect programs if program is changed or overridden. We have solved this problem with shell’s trap mechanism.
  • 西 竜志, 山本 慎一郎, 小西 正躬
    2004 年 124 巻 4 号 p. 1029-1035
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    The storage allocation planning problem in warehouse management is to determine the allocation of products to the storage space and intermediate operations for retrieving products so as to minimize the number of operations, and maximize the collected number of products for each customer when the sequence of requests for inlet and retrieval operations are given. In this paper, we propose an efficient beam search method for generating a near optimal solution with a reasonable computation time. A heuristic procedure is also proposed in order to reduce a search space in the beam search method by using the information of subsequent inlet and retrieving requests. The validity of the proposed method is confirmed by comparing the results with the optimal solution derived by solving an MILP problem. The effectiveness of the proposed method is demonstrated by solving an actual large-sized problem consisting of more than 3000 operations.
〈ソフトウェア・情報処理〉
研究開発レター
電子材料
電子・集積回路
  • Kei Eguchi, Fumio Ueno, Takahiro Inoue
    2004 年 124 巻 4 号 p. 1046-1047
    発行日: 2004年
    公開日: 2004/07/01
    ジャーナル フリー
    This paper introduces a family of ring-type switched-capacitor (SC) converters, which produces 0.75˜2.3 V outputs by being stepped-down from a 3.6 V input. The circuit consists of 2 series-connected DC-DC conversion blocks. Since the input voltage is converted twice by these blocks, the proposed circuit can provide more various types of output-voltages than conventional ring-type converters. Furthermore, the efficiency and the hardware-cost are improved since the number of driver circuits can be reduced. As an example, the circuit is designed with 17 types of output voltages in the range of 0.75˜2.3 V. The SPICE simulations show that the power efficiency is more than 80 % in the output current range from about 110 to 360 mA.
feedback
Top