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
117 巻, 8 号
選択された号の論文の26件中1~26を表示しています
  • 関根 慶太郎
    1997 年 117 巻 8 号 p. 993
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
  • 沖根 光夫, 勝原 範彰
    1997 年 117 巻 8 号 p. 994-999
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    Analog signal processing is very important technology for many reasons. Voltage followers are suitable active elements for a realization of high-frequency active RC circuits. A synthesis of chain active RC filters is proposed by which the lowpass function can be realized with voltage followers.
    This paper describes compensation method of the cutoff frequency shift for the active RC filter due to the low absolute accuracy of passive elements on a monolithic chip. The proposed filter is connected to a voltage controlled variable resistance, which determines resistance with a DC voltage Vc externally connected.
    As an example, a fifth-order Butterworth chain active RC filter is designed for frequency of 100kHz. SPICE simulation results are shown to demonstrate the effectiveness of the proposed method.
  • 杉本 泰博, 大原 健一郎, 飯田 哲也
    1997 年 117 巻 8 号 p. 1000-1007
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    The possibility of realizing a low-voltage and low-power audio-band Σ Δ modulator has been studied. The current-mode circuit technique has been adopted to enable the modulator operate in less than 3V of supply voltage. The modulator uses analog integrators instead of switched-capacitor-type integrators to avoid the use of analog switches, which enables low-voltage operation of the modulator circuit. A newly designed voltage-to-current converter is used to form an analog integrator. The designed circuit was actually fabricated by using CMOS 0.6μm process and its characteristic was evaluated. The signal-to-noise ratio of 68dB in 20kHz bandwidth for the Istorder modulator construction was measured from the 2.5V supply voltage when the oversampling ratio is 256. However, only 72dB of the signal-to-noise ratio and 65dB of distortion were measured for the 2nd-order modulator construction, although the operation of the oversampling ratio of 48 was successfully confirmed. We concluded that the low-voltage and low-power operation with less than 3V of supply voltage is possible for the audio-band Σ Δ modulator, although the signal-to-noise ratio and distortion issues in the 2nd-order configuration are left for future study.
  • 太矢 隆士, 吉田 聡, 山岡 信介, 松本 修一, 吉田 良和
    1997 年 117 巻 8 号 p. 1008-1014
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    In a large scale broadband communication system, thousands of high-speed serial data interconnections are used and a bit synchronization circuit (a clock and data recovery circuit) is required in each of the receiver side interconnection circuit. In this paper, the requirements and the implementation of a bit synchronization circuit for the interconnection are considered, and one solution is proposed. In the proposed circuit, the oscillation phase of a VCO is directly controlled by the trigger signal extracted from the input data. Synchronization capturing is quick and the circuit is applicable to a burst signal. The circuit tolerates jitter and phase variation of the incoming data. The circuit requires no external components, and is suitable for an integrated circuit. The circuit has been implemented using a 0.5um CMOS process and the data recovery operation from a 440Mbps pseudo-random pattern was confirmed. Data acquisition is accomplished within three clock periods from a 440Mbps burst data.
  • 李 裕次, 佐々木 圭則, 鶯塚 日出夫, 関根 好文
    1997 年 117 巻 8 号 p. 1015-1020
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    Recently the mobile communications, which are represented by the cellular phones, have been developed rapidly. A VCXO (Voltage Controlled Crystal Oscillator) is necessary to compose of a PLL and so on. The oscillation frequency of the VCXO is usually controlled by variation of capacitance of a varactor diode. In order to get wide variation of the frequency, it is necessary for the variable capacitance circuit of the VCXO to have wide variation, But, it is difficult to realize the varactor diode on an IC chip.
    First, we propose to apply a Miller capacitor to the variable capacitance circuit. As a result, it is verified that the variable capacitance circuit, which is operated with small amplitude, is so effective for the variable capacitance circuit of the wide variable range oscillator for IC.
    Next, we propose to apply this variable capacitance circuit to a crystal oscillator using a bipolar junction transistor. As a result, when the external capacitance of the MOSFET is about 10 [pF] and the capacitance ratio of the crystal resonator is 220, the VCXO using this variable capacitance circuit is shown to have wide variation of the frequency over 400 [ppm].
  • 松元 藤彦
    1997 年 117 巻 8 号 p. 1021-1027
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    This paper describes a high-performance continuous-time filter which is important in various electrical equipments. Firstly, a novel phase compensation technique for high-frequency integrators is proposed. The phase lead shift is realized by using a compensation capacitor. The ratio of the zero frequency and the pole frequency can be determined arbitrarily.
    Secondly, an elliptic low-pass filter using differential signal input integrators is discussed. A method to realize a high-precision elliptic filter is proposed. The effectiveness of the proposed techniques is confirmed by SPICE simulation. The simulation results of the integrator show that a flexible phase compensation has been achieved. The simulation results of the filter show that the elliptic filter with cutoff frequency of 10MHz can operate at a low power supply of 2V. The power dissipation is 4.72mW for a power supply of 2V.
  • Hironori Minamizaki, Takashi Taguchi, Tetsuro Itakura, Shoichi Iwamoto ...
    1997 年 117 巻 8 号 p. 1028-1034
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    Two types of low output offset 8-bit signal driver ICs for XGA/SVGA TFT-LCDs are described. These driver ICs generate 256 gray-scale levels using 11 reference voltages. One type incorporates a DAC using a capacitor array (C-DAC) and the other a DAC using a resistive divider(R-DAC). Using a fast offset-cancel circuit, the output-offset voltage of each driver has been reduced to within ±5mV (1/2 LSB) with the settling time of 10μsec for 150pF load capacitance. These driver ICs can operate at up to 55MHz at 3.3V. Both driver ICs have 309/300 outputs. They were fabricated using standard 0.6μm CMOS technology for logic ICs. The die areas of the drivers are 38.9mm2 for the C-DAC, type and 47.6mm2 for the R-DAC type; it is found that the C-DAC type is more cost-effective than the R-DAC type.
  • 石井 啓友, 高木 茂孝, 藤井 信生
    1997 年 117 巻 8 号 p. 1035-1042
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    Clock feedthrough (CFT) error is one of the most important problems for switched current (SI) circuits. This paper proposes a SI circuit which can reduce CFT error drastically. The proposed circuit will theoretically reduce both signal-dependent and independent errors by using CMOS switches under a fixed and appropriate bias. Although conventional circuits based on a similar idea need operational amplifiers or additional capacitors, the circuit proposed in this paper requires only MOSFETs. The proposed circuit can reduce clock feedthrough current with less power consumption and chip area compared to those of conventional circuits. An automatic tuning circuit, which controls the gate potential appropriately, is also proposed. Simulation results demonstrate the effectiveness of the proposed circuits.
  • 浦 智則, 高木 茂孝, 藤井 信生
    1997 年 117 巻 8 号 p. 1043-1052
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    A simulation system with low-level subcircuit simulators coexisting in the system level is proposed. In the top-down VLSI design strategy, a system-level simulation is indispensable not only to determine the system structure, but also to test the detailed design of a subcircuit in the whole system since the detailed design of an analog block is not exactly the same with its behavioral specification. Therefore, a system level simulator, where detailed design simulators, such as SPICE, coexist and synchronize together, is strongly desired. The proposed simulation system is designed based on the object-oriented scheme so that a new simulation algorithm or a signal conversion model can be easily introduced. Also an acceleration algorithm for Gauss-Seidel iterations is proposed to reduce the simulation time.
  • 板倉 哲朗
    1997 年 117 巻 8 号 p. 1053-1054
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
  • 小川 岳, 熊谷 寿彦, 鈴木 達也, 大熊 繁
    1997 年 117 巻 8 号 p. 1055-1061
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    In the motion or process control system, a variety of design techniques are proposed because of the demand of the high performance. The higer performance we demand, the higer the degree of the controller becomes. For practical use, these controllers are usually calculated by the digital controller. The controller is generally designed by a CAD system and is implemented with a microprocessor. But the microprocessor does not have enough precision to realize the result of the design by the CAD system. Therefore the system performance is deteriorated by the finite word length (FWL) effects. To cope with FWL problems, lots of design methods were considered in the siganl processing field and high-ordered digital filters are often used. Among these methods, the implementation technique based on the state-space realization can be minimize the sensitivity for the purturbation of coefficients. Noting the fact that the optimal realizations which mean same transfer function are unique only up to an orthogonal similarity transformation, we must chose the realization within this class of optimal realizations. In this paper, we show the algorithm to find a state-space realization which minimizes the frequency weighted sensitivity measure of the controller performance. Furthermore we show some experimental results to verify the effectiveness of the proposed algorithm.
  • 井野 秀一, 伊福部 達, 和田 親宗, 敦賀 健志, 泉 隆, 田中 敏明
    1997 年 117 巻 8 号 p. 1062-1068
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    Tactile information processing was studied based on psychophysical experiments in order to design tactile displays and sensors for tele-existence robots or virtual reality devices. The identification characteristics of the tactile sense concerned with the quality of materials were investigated when human fingertips touched the surface of various materials. In addition to obtaining the tactile characteristics, the temperature change of the skin surface of a fingertip was measured when the fingertip contacted the surface of materials. From these experimental results, the temperature difference from the normal skin temperature was found to be very important to recognize what kind of materials contacted the skin surface. Moreover, it was found that the identification characteristics concerning the quality of materials were different between passive touch and active touch.
  • 池内 光雄, 平澤 宏太郎, 大林 正直
    1997 年 117 巻 8 号 p. 1069-1075
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    Universal Learning Network (ULN) which can be used to model and control large scale complicated systems has been reported. The ULN is made up of nonlinearly operated nodes and multi-branches with arbitrary time delays between the nodes. Because of that, the ULN is capable of modeling in a natural way the large scale complicated systems which are hard to be modeled by the commonly used neural networks (NN). But, the ULN can't be applied to discrete event systems, because the ULN can only treat the system whose behavior is continuous.
    In this paper, a discrete event oriented learning network which is called Automaton Learning Network (ALN) is presented, which is also based on the architecture of the ULN. As an ALN is a discrete event system, leaning algorithm such as Back Propagation in NN, which is based on differential calculus can not be utilized for the ALN. Therefore a new local search optimization method which has the capability of intensification and diversification is proposed.This method is called RasVan which is an abbreviation of Random Search with Variable Neighborhood. One of the features of RasVan is that when there is quite a possibility of finding good solutions around the current one, intensified search for the vicinity of the current solution is carried out, on the other hand, when there is no possibility of finding good solutions, diversified search is executed in order to find good solutions in the region far from the current solution. Finally, simulations of a simple ALN which is composed of a controlled automaton and a control automaton are carried out in order to study the dependence on the initial states in the ALN and study the fundamental characteristics of RasVan.
  • 知識・規則を用いて
    前田 宏, 熊野 信太郎, 宮本 一正, 桑原 耕治, 古橋 武, 内川 嘉樹
    1997 年 117 巻 8 号 p. 1076-1083
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    For recognition of hallmarked characters on metal surfaces of various equipments, shadow image of characters by appropriate lighting is used. In actual environments where lighting conditions are limited, indistinct shadow images with noises make recognition difficult because segmentation of characters is not properly done.
    In this paper, a new approach is proposed to solve this problem. In addition to evaluating mutual correlation between image data of each target character and candidate template, dynamic programming method (DP matching) is applied to combine segmentation and recognition of characters. Prescribed rules and knowledges related to character strings are used for matching. Proposed approach is effective for not only constant pitch strings, but also constant gap strings.
    Experiments are done to show effectiveness and usefulness of this approach with respect to improvement of recognition rate under actual operational condition.
  • 前田 道治, 宮島 廣美, 村島 定行
    1997 年 117 巻 8 号 p. 1084-1091
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    Vector quantization is used for both storage and transmission of speech and image data, and is often required the algorithm that minimizes the distortion error. To obtain the minimum distortion error in neural networks for vector quantization, reformatory competitive learnings have been introduced. In a large number of algorithms, the self-creating neural network and self-deleting neural network are known for showing fine characters. In this paper, we improve the self-deleting neural network, and propose a generalization algorithm combining the creating and deleting neural networks. The algorithm is clarified as the following descriptions: At first, a few weight (reference) vectors are prepared, the algorithm is processed with the self-creating algorithm, and vectors are created automatically. Next, the algorithm is processed with the self-deleting algorithm, and weight vectors are deleted sequentially to the fixed number of them. Experimental results show the effectiveness of the proposed algorithm.
  • 鹿山 昌宏, 杉田 洋一, 諸岡 泰男, 福岡 昇平
    1997 年 117 巻 8 号 p. 1092-1100
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    One of major problems when we are going to construct a diagnosis system with actual plant data is difficulty in correcting sufficient number of fault data, while it is rather easy to obtain many data from normally working plant. To overcome this, we have investigated how to establish an effective diagnosis system only with normal plant data.
    In the previous paper, we proposed a distributed diagnosis system combining the Immune-type Network and Learning Vector Quantization for detecting fault sensor outputs accurately in industrial plants. The developed system can be created by training with normal data and detect abnormal sensor outputs by integrating local diagnosis results for pairs of sensors, which are obtained from similarity between diagnosis data and their nearest quantization vectors.
    This paper discusses how to improve matching capability at vector quantization spaces, which introduce local diagnosis results. First we clarify the effectiveness of the judgement whether the both sensor outputs are normal or not by matching a diagnosis vector and its several neighbor quantization vectors. Then fuzzy quantization vectors having their fuzzy function and consensus function created by superposing fuzzy functions are proposed to perform the mentioned matching execution simply. Lastly diagnosis capability of the developed system is evaluated using a proto-type system for detecting fault sensors of a reheating furnace plant. By introducing fuzzy quantization vectors, remarkable diagnosis capability is obtained compared with the diagnosis with conventional vector matching diagnosis.
  • 崔 康賢, 中村 秀男, 小林 尚登
    1997 年 117 巻 8 号 p. 1101-1106
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    This paper discusses a speed estimation problem of very low speed servo systems. Speed estimation can be easily done by conventional methods such as tachometer, observer, numerical differentiation and so on. However, speed estimation for very low speed systems still includes some engineering problems on precision. This paper proposes a simple way to derive a speed estimation by using the output difference in long duration and inputs sequence of the duration. We have also verified this method through an experiment of hydraulic servo system.
  • 村田 弘志, 原田 義治
    1997 年 117 巻 8 号 p. 1107-1112
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    Deadbeat control of the discrete-time systems has attracted particular interest because it ensures setting deviation response to zero within finite time, and suggests fast responsivity. By using controllers which include some delay-element, however, the deadbeat characteristics can be realized in the continuous-time domain and such a tracking system have been obtained.
    In this paper, we propose a new design method of continuous deadbeat control system without using delay-element. Our new controller is composed of a digital deadbeat compensator using state feedback and a series smoothing filter. Illustrative numerical examples are presented.
    The method proposed here is very simple comparing with other methods and it seems to useful for practical application.
  • 張 元勝, 大久保 重範
    1997 年 117 巻 8 号 p. 1113-1118
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    A model following control system (MFCS) can give general output signals following the desired ones. Since MFCS permits general signals as desired ones, its range of application is very wide. In previous studies, a method of nonlinear MFCS for continuous-time system was proposed by S. Okubo. In this paper, the method of nonlinear MFCS will be extended to discrete time nonlinear systems. Nonlinear systems which are dealt with in this paper have the property of norm constraint, _??_f(v(k))_??_ ≤ α+β_??_v(k)_??_γ, where α≥0, β≥0, 0≤γ≤1. When 0≤γ and γ<1, it is easy to extend the method to discrete time systems. But in the case γ=1 discrete time systems, the proof becomes difficult because the transfer function from f(v(k)) to v(k) can't be a positive real function. In this case, to ensure that internal states are stable, a new criterion is proposed.
  • 変調方式と安定度の評価
    渡部 博道, 東 秀樹, 中野 博之, 佐藤 孝, 大河 正志, 丸山 武男, 榛葉 實
    1997 年 117 巻 8 号 p. 1119-1126
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    Frequency stability of a semiconductor laser is deteriorated under direct frequency shift keying (FSK) when the stabilization method uses an external frequency reference in a normal way. Therefore, the PEAK method was devised for the purpose of improving frequency stability under direct FSK. The PEAK method, however, requires two frequency components in principle. Therefore, the PEAK method does not work effectively in a given communication system, which produces a succession of identical frequency components occasionally in a normal FSK method. In this paper, two modulation methods are used and confirmed its operation to make the PEAK method effective. The evaluation of the frequency stability uses the beat note between two stabilized laser beams, signal and reference lasers, here. The reference laser is stabilized by the method using the magneto-optical effect. This reference laser reduces the frequency broadening of the beat note and makes it possible to choose the beat frequency. The obtained frequency stability is improved about one digit by using the PEAK method.
  • 小野 俊之, 加賀美 晃, 小坂 満隆
    1997 年 117 巻 8 号 p. 1127-1133
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    For the management of a distribution center, efficiency of picking ordered goods is very important. The efficiency of an automatic picking system is decided by its system structure and its operation method. Conventionally, to improve the operation efficiency, methods about order processing sequence and goods allocation to a rack were researched. However, there's no consideration of goods allocation to plural racks, or changes of goods allocation are needed while picking. In this paper, we propose an operation method for the automatic picking system, which is composed of plural sets of a picking-hand and a rack, and the method is especially suitable for recent piece-picking in which several kinds of goods of different lacks should be collected into one case by an order. The proposed method is a combination of following: (1) the method of allocating goods to lacks so that each picking-hand picks equal quantity of goods for orders, and (2) the method of deciding order processing sequence so that orders processing time's variance among picking-hands is small. Finally, it is shown by simulations that the proposed method reduces the required time by about 16% compared with a simple method.
  • 柴田 直人, 朝日 昇, 雨宮 好仁
    1997 年 117 巻 8 号 p. 1134-1139
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    We propose a new method for solving mathematical problems by using neural networks. This method is a kind of analog computation and is quite different from conventional solving methods, so it is expected that this method can solve problems that are intractable for conventional methods.
    The procedure for solution is as follows. We first prepare the circuit elements such as: function generators constructed by neural networks, differentiators, integrators, multipliers and adders. A problem equation is then implemented by using these circuit elements. A function signal is generated by the neural networks and the output of the problem circuit is observed, and the neural networks are trained so that the output will approache zero. If the training is completed, then we can obtain the solution to the problem. We simulated the method on differential equations and variation problems.
  • Myung-Mook Han, Shoji Tatsumi, Takaaki Okumoto
    1997 年 117 巻 8 号 p. 1140-1151
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    In this paper, we propose the use of Genetic Algorithm (GA) as a key element in the design of robust learning system. We describe a GA-based concept learner called GAIL, capable of inducing descriptive concepts in conjunctive statements. A motivation to use a GA is that GAs are able to explore large domains because they make use of a multipoint search strategy and are suitable to exploit massive parallelism. By incorporating a GA with a traditional method, the system has several important features. First, the system is robust. Second, the system introduces the modified evaluation criteria to evaluate the quality of a clustering. Finally, the system searches for an individual which refers to all the concepts at the same time. The effectiveness of this approach is illustrated by two exemplary problems and the clusters obtained with this system correspond well to categorization satisfied by people.
  • 山根 健治, 田中 正吾
    1997 年 117 巻 8 号 p. 1152-1158
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    Using acoustic sensors, the authors previously proposed an effective detection system of anomalous objects in tunnel excavation with shield tunneling machines to prevent the machine from getting damages. Not many observation data being available in the detection, optimal observation policy is important. From this viewpoint, the authors next considered an optimal sensor location on the cutter plane of the machine and also the optimal observation policy, for the case where three receiving transducers were used to detect lump objects. The optimal location of the receiving transducers was shown to be given as arbitrary equally-spaced points on the outer circle of the cutter plane. The optimal rotating angles were also found to be arbitrary. In applications, however, it is often difficult to locate sensors at arbitrary positions of the cutter plane. This paper thus considers the optimal observation policy for detecting anomlous lump objects effectively under the practical constraint that the sensors are located only along a diameter of the cutter plane.
  • 水本 巌, 益子 信郎
    1997 年 117 巻 8 号 p. 1159-1162
    発行日: 1997/07/20
    公開日: 2008/12/19
    ジャーナル フリー
    A low noise 8 channel spectroscopic measurement system for 1-2μm near infrared region was developed. Using commercially available InGaAs PIN photodiode with charge integrating amplifier, the minimum detectable power of 10-14 W were achieved in all channels. The simultaneous measurement wavelength range of the system was from 1060 to 1550 nm. The spectral resolution of the system was 70 nm at 1 channel. This is the first near infrared spectroscopic polychromator using the multi-channel ultra low nosise array detector.
    We demonstrate the measurement of the spectrum of singlet oxygen using this system.
  • 1997 年 117 巻 8 号 p. 1163
    発行日: 1997年
    公開日: 2008/12/19
    ジャーナル フリー
feedback
Top