計測自動制御学会論文集
Online ISSN : 1883-8189
Print ISSN : 0453-4654
ISSN-L : 0453-4654
34 巻, 10 号
選択された号の論文の36件中1~36を表示しています
  • 高橋 宏, 黒田 浩一
    1998 年 34 巻 10 号 p. 1271-1279
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    This research focused on a vehicle control system capable of selecting suitable control parameters based on detection of the driving environment by multiple sensory devices, including video cameras and laser radar. An information space mapping type of sensor fusion system is proposed whereby multiple sensor signals are represented as vectors in sensor signal space, and the system variables needed for vehicle control from these vectors are mapped to corresponding vectors in control parameter space.
    A hierarchical fuzzy integral model, represented in terms of declarative relationships, is adopted for the process of variable mapping, thereby avoiding the laborious complexities of sensor fusion design. Moreover, the driving environment recognition characteristics of individual drivers are treated in terms of perceptive sensibilities relative to affordance, making it possible to vary the lambda-fuzzy measure according to personal driving behavior. The applicability of the proposed system to vehicle control has been evaluated and fundamentally confirmed on the basis of simulations.
  • 宮崎 早苗, 中川 透
    1998 年 34 巻 10 号 p. 1280-1288
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    In the near future that the various satellite image data will be obtained freely, new utilization of remote sensing data may be derived from new image analysis methods that are robust against light condition. The ground reflectance is one of the robust value against light condition which presents the feature of the ground. We have developed a ground reflectance estimation method from satellite image data and topographical information by monte-calro simulation based on the radiative transfer theory, and then we have applied our new method for two different Landsat TM images, and tried to classify the ground for its land cover using the estimated reflectance of each pixel. As a result, we found that the estimated reflectance was robuster than Digital Number of each pixel, and the classification results using reflectance had only a few influence of light condition compared with that using Digital Number.
  • 小足 克衛, 重田 修作, 横田 博, 木村 雅昭
    1998 年 34 巻 10 号 p. 1289-1295
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper describes the methods and apparatus for measuring accurately the thickness of thin oriented polymer films. Three key aspects of the new method are as follows:
    (1) The technique eliminating optical interference effects occurred with thin films by using plane polarized light incident at Brewster's angle;
    (2) The implementation of multivariate regression models based on the inverse representation of Beer's law or P-matrix approach with optimal wavelengths, canceling dichroic effects with anisotropic materials such as oriented polymer films;
    (3) The design of optics to be quite tolerant to movement of the film in both height and angle;
  • 吉田 浩隆, 米田 克哉, 鉄矢 仁
    1998 年 34 巻 10 号 p. 1296-1303
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    The usual way to set up a telegraph pole is to use an earth auger machine. However, since the locations of underground utility pipes and cables often are not accurately recorded, buried objects are often damaged during drilling. Although electromagnetic induction and electromagnetic waves have been used for sensing underground objects non-destructively, the former cannot be used for non-metallic objects, and the latter produces radar images that are difficult to interpret.
    We have studied a method of sensing underground objects while drilling in earth based on vibration analysis. In order to measure the vibration signal, we attached a toothed cylindrical casing with an acceleration sensor mounted inside to the tip of the auger drill. The results obtained in this studies are as follows:
    1) Underground objects could be detected by observing the change in the acceleration signal upon contact.
    2) Buried pipes could be distinguished from other objects (e.g. stones and pebbles) by analyzing the spectrum of the acceleration signal.
    3) We have developed a underground object sensing system for earth auger machine. And it has been confirmed that the system is effective for actual field use.
  • Harumi KAMADA, Nobuharu AOSHIMA
    1998 年 34 巻 10 号 p. 1304-1309
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    We propose an analog Gabor transform filter. Gabor function is widely used for digital time-frequency analysis as non-orthogonal wavelet. However, it has never been reported in analog filtering. The proposed filter is realized as cascade connection of CFOS (complex first order system). The impulse response of the filter will be considered as the product of a window function and a sinusoidal wave. Because the window function is approximated to Gaussian function, we regard the whole system as Gabor transformer. We have actually constructed the proposed filter, and measured the characteristics. Its mathematical foundation and experimental results are presented.
  • 小畠 時彦, 大岩 彰
    1998 年 34 巻 10 号 p. 1310-1315
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    In order to generate precise dynamic pressure, we are developing a dynamic pressure generator using a newly rotating valve. This paper describes the principle of the valve and the generator and some experimental results. The generated waveform is square whose amplitude and fundamental frequency can be changed. When the supplied difference pressure was 5kPa and the fundamental frequency was changed from 1 to 100Hz, we measured the waveform by a pressure transducer, and evaluated the amplitude and the mean value of the waveform. As an experiment result, the amplitude and the mean value of the generated pressure were almost constant in the frequency range from 1Hz to 50Hz.
  • 湊 淳
    1998 年 34 巻 10 号 p. 1316-1320
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    New laser radar technique using correlation of chaotic modulation is presented. Chaotically modulated laser beam is transmitted into the air. Return signal and modulated laser power are digitized and recorded. Because correlation of chaotic signals behaves like a delta function, the correlation function of return signal and monitored power corresponds to the return signal of conventional laser radar using pulse laser. Numerical experiment was carried out and validity of the technique was confirmed.
  • 出口 光一郎
    1998 年 34 巻 10 号 p. 1321-1328
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper describes a factorization-based algorithm that reconstructs 3D object structure as well as motion from a set of multiple uncalibrated perspective images. The factorization method introduced by Tomasi-Kanade is believed to be applicable under the assumption of linear approximations of imaging system. In this paper we describe that the method can be extended to the case of truly perspective images if projective depths are recovered. We established this fact by interpreting their purely mathematical theory in terms of the projective geometry of the imaging system and, thereby, giving physical meanings to the parameters involved. We also provide a method to recover them using the fundamental matrices and epipoles estimated from pairs of images in the image set. Our method is applicable for general cases where the images are not taken by a single moving camera but by different cameras having individual camera parameters. The experimental results clearly demonstrate the feasibility of the proposed method.
  • 佐々木 公男, 平田 克己
    1998 年 34 巻 10 号 p. 1329-1337
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    For the purpose of developing refined devices for environmental recognition in robotics, a practical method for 3D-localization of a stationary random acoustical source in near-field is proposed by newly introducing 3 fixed-detectors. By using two pairs of estimated auto-and cross-spectra of signals detected simultaneously at 3 fixed detectors, two sets of information about difference in propagation distances and the amplitude attenuation of wave propagation are derived in such a manner that the derivation is independent of the shape of power spectra of detected random signals so as to be applicable to any stationary random acoustical source and additive corrupting noises at detectors are cancelled out as completely as possible. Then, by combining these information with the geometric relation of signal detection, the source position in 3D-space is estimated. After the principle of the proposed 3D-localization and the required conditions being made clear, 3D-distribution area of the position estimate is theoretically evaluated as concentration ellipsoid, based on the sampling variations of estimated spectra, and a simple method of evaluating the 3D-distribution or localizable area with prescribed relative errors is derived from the maximum length of major axes of the ellipsoid. To show the effectiveness of the theoretical results and make clear the influence of the source range, SNR, at detection and spatial distance between detectors on the final source distribution or localizable area, numerical analyses are carried out by changing them parametrically, the result of which illustrate the effectiveness as well as fundamental characteristics of the proposed method under practical circumstances.
  • 小林 正信, 坂手 弘明, 佐久間 史洋, 小野 晃
    1998 年 34 巻 10 号 p. 1338-1344
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    A multiwavelength radiation thermometer having a hybrid detector of a 64-element silicon-germanium linear array and an infrared detector of a 48-element InSb linear array was developed for measurements of temperature and emissivity. A wide wavelength range detection by one fixed grating from 0.55μm to 1.6μm were realized by the hybrid detector. The InSb detector and a LiF prism were used for the long wavelength range from 1.3μm to 5.3μm.
    The two-detector optical system without mechanical wavelength scanning and light chopping improved the stability and the response time as short as 256μs to scan a total of 112 wavelengths. In a continuous 6-month operation without optical adjustment, no wavelength drift and 5% reduction in responsivity were observed.
    A reference radiation source of a plane blackbody furnace was used for the compensation of the responsivity reduction and the atmospheric absorption of the optical pass of the multiwavelength radiation thermometer.
    The temperature scale and the wavelength of each detector element were calculated from calibration data. Two blackbody furnaces and 24 narrow-band-pass filters were used in the calibration. The noise equivalent. temperature differences were estimated from the temperature scales and noise data of the analog signals when the time constant of the amplifiers was 100ms and the target was a blackbody operated at 1052°C; they were 0.16°C, 0.08°C, 0.24°C, and 0.54°C at the wavelengths of 0.55μm, 0.66μm, 2.0μm, and 5.0μm, respectively.
  • 初澤 毅, 井上 尚紀, 早瀬 仁則, 坂野 憲幾
    1998 年 34 巻 10 号 p. 1345-1348
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    A surface roughness reference has been developed by using an anisotoropic etching of a (100) silicon wafer. As the etching process of the reference is dominated by the crystal orientation, it has a geometrically accurate profile compared with those manufactured by conventional mechanical tooling techniques. The reference has a serial triangle profile so that calibration in both horizontal and vertical dimensions can be performed easily and simultaneously. Dimensions of 30 to 3μm-pitch references are evaluated by using a stylus profiler, resulting in sub-μm pitch errors and an average valley depth error less than 0.32μm. Little damage by the stylus is observed after measurements, since the silicon is harder than metal materials.
  • 大山 真司, 曹 麗, 小林 彬
    1998 年 34 巻 10 号 p. 1349-1356
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper describes a methodology for obtaining a higher precision for a position measurement in an enclosed signal field by introducing the concept of a zooming functional measurement. First, the way of thinking from a measurement equation to the zooming functional measurement was shown. This shows that there exists the effect of reducing the accidental errors in the case of intentionally changing the sensitivity but not changing the amount of noise in the active sensing system such as an enclosed signal field. Next, this method was applied to the position measurement by the enclosed signal field. New variable parameters that included both initial phases and brightness ratios were considered in the conventional lighting functions. The proper adjusting of these parameters makes it possible to get a desired sensitivity at intended positions. By using this operation, the zooming functional two-dimensional position measurement could be performed. Finally, these effects were experimentally confirmed by a measurement system with four linear LED arrays and a photo sensor. In the experiments, the setting sensitivity was changed from 1 to 16, and the results showed that the standard deviation at sensitivity 8 was 1/5 of that for 1. This fact revealed that the accidental error was reduced by this zooming function.
  • 増田 士朗, 岡本 太, 井上 昭
    1998 年 34 巻 10 号 p. 1357-1365
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    In this paper, we propose a new design scheme of hybrid adaptive control system using a fixed compensator for single-input, single-output, continuous time invariant plant. In the proposed method, adaptive adjusting law can be simplified because parameters are updated by using estimated error at arbitrary sampling instants. Furthermore, boundedness of all the signal in the closed system is proved. In general, the stability of adaptive systems cannot be assured by only using sampled estimated error. However, in this paper, the difficulty is overcome by utilizing the fixed Compensator. Finally, a simulation result is illustrated in order to ensure the effectiveness of the proposed method.
  • 佐野 滋則, 尾形 和哉, 藤井 省三, 早川 義一
    1998 年 34 巻 10 号 p. 1366-1373
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    In recent years, many approaches to make a model using the system identification method in order to design the robust control system are proposed. In these approaches, there are methods using the set membership identification technique. In the set membership identification, the model parameters are assumed to be time invariant. Therefore, it is difficult for the set membership identification to deal with an identified object whose parameters are time variant. In this paper, in order to identify time variant model parameter, the set membership method is developed such that the identified model can explain the observed input-output relation using the time-variant parameters which exist in the identified parameter set. The optimization problems are proposed in the case that the parameter of the nominal model is known and unknown.
  • 内山 剛志, 今井 純, 和田 清
    1998 年 34 巻 10 号 p. 1374-1379
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    In this paper, robust identification of discrete time systems in time domain is studied. An approach is proposed to estimate the nominal model, minimizing the errors, with iterative computation on numerator and donominator. Methods are proposed to reduce the influence of noise on the proposed scheme. Numerical study shows effectiveness of these methods for idenifying characteristics in low frequency range of systems in the presence of noise.
  • 安田 一則, 能宗 文子
    1998 年 34 巻 10 号 p. 1380-1387
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper is concerned with a problem of robust stabilization with disturbance attenuation for a descriptor system with norm bounded uncertainties. The uncertainties may lie in not only the state matrix and the input matrix but the coefficient matrix for the derivative of the descriptor variable. We propose a sufficient condition under which the system is robustly stabilizable with disturbance attenuation, and present a static descriptor variable feedback gain and dynamic output feedback controllers for such a stabilization.
  • 申 鉄龍, 武藤 康彦, 田村 捷利
    1998 年 34 巻 10 号 p. 1388-1394
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper presents an approach to design adaptive controller for nonlinear system with modelling errors. The modelling errors are described by nonlinear smooth perturbated function in the state. Robust regulation and robust disturbance attenuation are considered by using adaptive techniques. It is shown that if zero dynamics of the system is robust stable, then one can construct the Lyapunov function for the overall closed loop system by using the Lyapunov-based recursive design method.
  • 十河 拓也, 足立 紀彦
    1998 年 34 巻 10 号 p. 1395-1403
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    If a linear continuous-time system is considered on a finite time domain, there is no diverging variable of the inverse systems even if the transfer functions have unstable zeros. However, if a sampled-data systems is discussed on the finite time domain, it is not a trivial matter to determine whether there is diverging variables or not because shrinking the sampling period on the finite time interval implies increasing the number of sampling points. There is a possibility that the increase of the time interval for the discrete-time system may cause diversion. In this paper, we show that there is no diverging variable of the inverse of sampled-data systems with 0-order holders if the relative degree of the continuous-time transfer function is 0 or 1. It is also demonstrated that such a property is independent of stability of zeros. The main result implies that we can formulate a problem of digital iterative learning control as a minimization problem of output errors on the sampling points.
  • 土井 智晴, 大須賀 公一, 小野 敏郎, 川谷 亮治
    1998 年 34 巻 10 号 p. 1404-1410
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    Previous results concerned with a disturbance observer have disadvantages as follows: (i) Discussion about robust stabilization for the closed loop is not enough. (ii) Operation for structured time-varying uncertainties is not correct. First, we show the relation between H control with structured uncertainties and disturbance observer. Then, we show that a design method of a disturbance observer with a quadratically stabilization is same as the problem of quadratic stabilization of systems with parameter uncertainty via output feedback. Next, we consider the robustness for previous and this method of disturbance observers. Finally, with a simulation for an electromagnetic suspension system, we show that the design method has effectiveness.
  • 金 英福, 池田 雅夫, 藤崎 泰正, 小林 真樹
    1998 年 34 巻 10 号 p. 1411-1418
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    Robust stability of a two-degree-of-freedom servosystem incorporating an observer is considered. A class of uncertainties allowed in the plant model is obtained, to which the servosystem is robustly stable for any gain of the integral compensator. The gain can be tuned arbitrarily to improve the transient behavior. Using the singular perturbation approach, the transient behavior attained by the limit of high-gain integral compensation is calculated. An example is presented, which demonstrates that the tracking response becomes faster when the integral gain is made larger.
  • 杉本 謙二
    1998 年 34 巻 10 号 p. 1419-1424
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper proposes a design method for quadratic stabilization of servo systems with partial pole placement. The method parametrizes all state feedback gains placing part of the poles at prespecified points for a nominal system augmented with an integrator. Then, the parameter is taken large enough to guarantee quadratic stability of the feedback system for a given uncertainty. This is possible under a certain sufficient condition concerning an H norm determined by the partial pole specification. Examples are given to illustrate the method.
  • 板宮 敬悦, 鈴木 良昭, 鈴木 隆
    1998 年 34 巻 10 号 p. 1425-1433
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper presents a robust scheme for the direct type of model reference adaptive control system (MRACS) and its stability analysis in the presence of unmodeled dynamics and measurement noise. This scheme is characterized by the use of adaptive law with dead zone and integral type of fixed compensator. The adaptive law prevents the adjustable parameters from drifting and the fixed compensator improves the deteriorated control performance caused by the use of dead zone. The stability of the system is analyzed by a new method which evaluates an exponentially weighted norm, called as L norm, of the regressor signals and then applies the Gronwall lemma. This method enables us to analyze the stability based on the L2-property of the parameter adjusting rate, which is guaranteed by the adaptve law. Effectiveness of the scheme is demonstrated by the simulation carried out for a simple plant with 1st order nominal part.
  • Jinwon KIM, Heung Won PARK, Chan Gook PARK, Jang Gyu LEE
    1998 年 34 巻 10 号 p. 1434-1439
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    An observability analysis for the two-position alignment of strapdown inertial navigation system (SDINS) is performed using an analytic approach utilizing the nonsingular condition of the determinant of a simplified stripped observability matrix. It is then confirmed by employing a numerical calculation of the error covariance propagation using a twelve-state stationary error model of SDINS including a vertical channel. Based on the results, completely observable conditions and degree of observability are introduced for the two-position alignment.
  • 加島 正, 石動 善久
    1998 年 34 巻 10 号 p. 1440-1447
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    Humans select one of infinite number of trajectories in voluntary arm movements and it is reported that the trajectory is rationally determined from dynamical and physiological characteristics of a human arm. Accordingly, it is necessary for human arm trajectory formulation to obtain a dynamic model and an objective function in which physiological characteristics of muscle is considered.
    This paper proposes a dynamic model and an objective function based on characteristics of muscle viscoelasticitiy and energy expenditure. Experiments have been conducted for human arm motions in a vertical plane and the trajectories given from the experiments and proposed trajectory formation are analyzed. As a result, formulated trajectories show good agreements with measured ones and it is verified that the trajectories which humans select have physiological advantages of minimizing energy expenditure in muscles. Furthermore, different types of trajectories are observed, which vary with humans in relatively large movements. In the proposed arm model, adjusting factors in the case of similar arm dimensions are only coefficients on muscle viscoelasticity and variety of trajectories according to humans are discussed to be caused by differences of muscle viscoelasticity characteristics.
  • 中山 万希志, 前田 知幸, 北辻 佳憲, 馬野 元秀, 井口 征士
    1998 年 34 巻 10 号 p. 1448-1456
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    In this paper, we discuss the fuzzy visual feedback control system for swirling flow melting furnace based on image processing and fuzzy theory. This system has following two blocks, one is the combustion evaluation block which evaluates the combustion condition in the furnace and the other is the manipulation value inference block which generates the manipulation value combining above combustion evaluation and temperature data in the furnace wall. In the evaluation block, we suggested to convert the physical feature extracted by image processing to the operator language which the control operator use to express the combustion condition from the slag flow image through the CCD camera. We define the two operator language, one is the conceptual operator language which is the whole expression for the combustion, for example “Good”, “Bad” “Usual” and etc., the other is the concrete operator language which is the concrete expression explaining the conceptual operator language for example “flow smoothly” “flow very well” and etc. In order to obtain the conceptual operator language from the image through the CCD, the engineers construct the image processing according to the concrete operator language, and the membership function for the physical feature extracted above processing based on the hearing and questionnaire for the operator. We named the results applying this membership function “intermediate operator language” related the concrete operator language. Finally, the conceptual operator language is obtained the fuzzy decision tree constructed by fuzzy ID3 using above intermediate language. Through the experiments, we confirmed this system is effective.
  • 川上 肇, 黒江 康明, 森 武宏
    1998 年 34 巻 10 号 p. 1457-1465
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper describes a new method by which images can be recognized robustly in the presence of random perturbations to the images. In general, a computer could recognize images by the following procedure: (1) Calculate a cross-correlation coefficient between an unknown image and each of images memorized as standards. (2) Find the memorized standard image which maximizes a maximum value of the cross-correlation coefficient. (3) Recognize that the input image belongs to a class of the found standard image. Although this conventional method is effective if imaging environments are calibrated, it does not always work well under a natural environment in the presence of random perturbations; therefore, a problem to be solved next is how to construct a method for the robust image recognition under the natural environment.
    To solve the problem in case that input images have a texture region constructed by a partial image i.e. a texel, we introduce, as a measure suitable for the robust image recognition, local point symmetry inherent in the cross-correlation coefficient between the input image and the partial image. We show that a disorder measure for the local point symmetry of the cross-correlation coefficient is robust against random perturbations; then, we propose a method for robust image recognition with the disorder measure.
    We have made experiments on recognizing input images distorted with random perturbations. Their results have shown that the input images can be recognized by our method more robustly than those done by the conventional method.
  • 兼田 昌子, 川田 明, 林 重雄, 徳井 一雄
    1998 年 34 巻 10 号 p. 1466-1471
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    The rope tester for elevator wire rope detects pulse signal induced by magnetic leakage flux from defected part, such as breaking of strand. At the same time, noise is induced by variation of magnetic leakage flux, due to vibration of wire rope.
    In this paper, the data processing method using wavelet transform is proposed to detect pulse signal from the noisy data. The characteristics of pulse signal from defected part and noise are made clear by analyzing measured data using multiresolution decomposition based on Haar basis. The threshold value for diagnosis is set up based on these results and diagnosed if the wire rope has defect part or not by the threshold value. The effectiveness of the proposed method is confirmed by comparing the results of diagnosis by this method with the result of visual check of inspector. It is expected to replace the data processing work by visual check with the automatic processing.
  • 大倉 典子, 前田 太郎, 舘 〓
    1998 年 34 巻 10 号 p. 1472-1477
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    For binocular visual space, it is well-known that horizontal lines which appear straight on a subjective frontoparallel plane are not always straight in the physical sense, and that the form of the lines depends on the distance from the observer. A similar phenomenon is also known for tactile space. Then, we conducted fundamental experiments on sound localization with distance, and confirmed the similar phenomenon for perceived binaural auditory space. On the other hand, we proposed a mathematical model for space perception to explain visual and tactile horopter curves by using biological information and constraints. this model, named ISLES (Independent Scalar Learning Element Summation) model, succeeded to explain the dependency on the distance of the horopter curves uniformly. From the results of other experiments we conducted, sound intensity and interaural time differences seem to be quite important cues for sound localization with distance. In this paper, these cues are employed instead of biological information to construct ISLES model to explain auditory horopter curves. The constructed model for space perception is examined by simulation. As the results of the simulation, the followings are obtained.
    •By using sound intensity and interaural time differences as parameters, the constructed model can explain the dependency of auditory horopter curves on distance.
    •The differences of the curves between subjects are explained by the differences of individual learning areas for space perception.
    •Relatively near part of the learning area of ISLES model for visual horopter is not included in the learning area of ISLES model for auditory horopter.
    That is, this model successfully explains the results of fundamental experiments on sound localization with distance.
  • 湯浅 哲也, 星野 充紀, 渡部 裕輝, 冨樫 整, 尾形 健明, 赤塚 孝雄, 鎌田 仁
    1998 年 34 巻 10 号 p. 1478-1486
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    Abstract. Electron spin resonance (ESR) spectroscopy is the only direct detective method of radicals which are substances with unpaired electrons. Recent studies in biochemistry have clarified that various kinds of radicals have relation to several biological phenomena, e.g., carcinogenesis, aging, biophylaxis etc. Therefore, in vivo ESR-imaging of spatio-temporal behavior of radicals has significance for pathology and physiology. So far, we have developed an ESR-CT algorithm based on deconvolution procedure, and developed an interactive data processing system for actual observation. Using this system, satisfactory in vivo CT images have been obtained from the viewpoint of morphological imaging. However, this method has the limitation for quantitative and dynamic analysis of bio-radicals. In this paper, we elucidate the reason why the conventional ESR-CT algorithm based on deconvolution is not enough for quantitative measurements, and propose a new algorithm to estimate projections, more quantitatively, based on modeling of ESR signals with Gram-Charlier expansion. Furthermore, we demonstrate its effectiveness by applying it to actual data from physicalphantom experiment and in vivo experiment for mouse's liver.
  • 永田 和之, 慶野 知治, 小俣 透
    1998 年 34 巻 10 号 p. 1487-1493
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper describes a manipulation system which can acquire a model of an object while manipulating it with a multifingered hand. We assume that each finger of the hand has a 6-axis force-torque sensor at its fingertip and that the object to be manipulated is a polyhedron. When the finger is in contact with a face of the object, the position and normal of the contact point can be calculated by using the data from the force-torque sensor. These parameters are then used to determine the plane of the contact face. The proposed method acquires the object model by intersecting the determined planes and incrementally refines it by manipulating the object. During manipulation, the proposed method checks if the fingers have moved to an unexplored face of the object. If they do, a new surface plane is obtained which is then intersected with the current object model. The acquired object model can be used for manipulating the object with the hand. Experimental results using a four fingered hand are shown to demonstrate the effectiveness of the proposed method. This method enables a multifingered hand to manipulate various objects of unknown shapes.
  • 星野 力, 湊 宗篤
    1998 年 34 巻 10 号 p. 1494-1500
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    An engineering problem associated with the emergent property of hierarchical systems is whether local behaviors in the lower level system can achieve a desirable coordination in the upper level system. The lower level system supposed in this paper is a car navigation system that pursuits egoistically (i.e. only for that particular system) the shortest-time routing in the network of roads. Even the car navigation system is egoistic, such a strategy is feasible by the use of the genetic algorithm(GA) that adapts to the global traffic condition and even achieves the coordination.
    The navigation system proposed here introduces the capability of adapting its routing strategy to the dynamically changing traffic condition. Each system has a set of weight associated with several routing policies, and the decision is made, at every branching point of the road, based on the weighted traffic information. GA algorithm selects, mutates and cross-overs the weight vectors in a population of navigation systems.
    In the simulation studies on typical road maps with simple and complex topologies, it has been demonstrated that the adaptive strategy of car navigation systems successfully adapt themselves to the interacting condition, and even solve an accidentally happening traffic jam.
  • Ru LAI, Fujio OHKAWA
    1998 年 34 巻 10 号 p. 1501-1503
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper presents a new discrete-time robot model which has very simple structure and independent to nonlinear force terms and external force term. Both pole-placement control scheme and adaptive control scheme are designed and implemented in controlling a 2-link manipulator disturbed by unknown external forces.
  • 江頭 成人, 中村 政俊, 久良 修郭
    1998 年 34 巻 10 号 p. 1504-1506
    発行日: 1998/10/31
    公開日: 2009/03/27
    ジャーナル フリー
    Transitional velocity ripples at each reference input time interval of mechatronic servo system were analysed by introduing an appropriate analytical model. The analytical result of the velocity ripples was evaluated based on an experiment of an industrial DC servo motor. The results gave important information for designing a controller of the mechatronic servo system for given required accuracy.
  • 1998 年 34 巻 10 号 p. e1
    発行日: 1998年
    公開日: 2009/03/27
    ジャーナル フリー
  • 1998 年 34 巻 10 号 p. e2
    発行日: 1998年
    公開日: 2009/03/27
    ジャーナル フリー
  • 1998 年 34 巻 10 号 p. e3
    発行日: 1998年
    公開日: 2009/03/27
    ジャーナル フリー
feedback
Top