All issues
Volume 36, Issue 11
Displaying 1-25 of 25 articles from this issue
- |<
- <
- 1
- >
- >|
-
Masami IWATSUKI, Koji SUZUKI, Takehiro YAMAGUCHI2000 Volume 36 Issue 11 Pages 887-893
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper proposes a new stereo vision method that can reconstruct 3D structures robustly by matching left and right images based on regions given by color segmentation. The proposed method enables correct matching of the regions by segmenting simultaneously stereo images into color regions and matching feature points on the edges of the same color regions. Furthermore, this method can reduce fault correspondences because the difference between segmentation results of right and left images is absorbed by carring out region matching using hierarchical color segmentation whose threshold to merge classes of color is changed step by step in order to obtain appropriate region correspondences independently of a select threshold value. This method can also detect and modify occluded parts by assuming that artificial objects consist of flat color regions approximately.View full abstractDownload PDF (3929K) -
Kajiro WATANABE, Harumi WATANABE2000 Volume 36 Issue 11 Pages 894-900
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper describes experimental investigations on a new vital bio-measurement method of the pneumatic and/or air mattress type. The method indirectly detects the vital bio-information such as the heart beating, respiration, body movement, coughing and snoring via the change in the pressure of the air in the air mattress on which a human stays and to which his vital movements influence. The blood pressure measurement is a special example of the pneumatic methods. The air mattress method requires nothing to him except that he must be on the air mattress. Thus the method is less restrictive than the blood pressure measurement and any conventional polygraph methods which stick the electrode to human body.
This paper treats practical problems, i.e., reduction of noise, human feeling to the air mattress, application of the method to commercially available air cushions or mattresses.
As the results, this method can be robust to the pressure noises in daily life by properly selecting the cut-off frequency of the filters. Further the method works under a very thin air mattresses such that the human can not notice their existence as well as to the variety of existing air cushions and mattresses. The air mattress method is most stable to measure the vital bio-information with the least restriction.View full abstractDownload PDF (5128K) -
Toshiaki YONEZAWA, Shogo TANAKA2000 Volume 36 Issue 11 Pages 901-908
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSFor an accurate on-line measurement of ship's attitude, the authors previously developed a measurement system with one servo-type accelerometer and two servo-type inclinometers which were appropriately located on the ship. The system realized a higher accurate on-line measurement using an effective update algorithm for the candidates on system unknown parameters. The system, however, assumed an exact knowledge on the location of the rolling axis. This caused much degradation in the measurement accuracy in the case of unknown rolling axis. Such a case is, what is worse, usual practically.
This paper proposes a sensing system which estimates the ship's attitude despite no information on the location of the rolling axis.View full abstractDownload PDF (1354K) -
Koichi SAGAWA, Yutaka SATO, Hikaru INOOKA2000 Volume 36 Issue 11 Pages 909-915
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSNon-restricted and precise measurement of horizontal displacement during walk is discussed. Walking velocity is derived by the integration of horizontal acceleration which is measured by a sensor system attached to the subject's toe. The walking distance results from the integration of the walking velocity. The sensor system consists of a three dimensional accelerometer and a ceramic gyro. The accelerometer measures three dimensional acceleration of the toe and the gyro senses the angular velocity of pitching angle of the sensor system. The axis of the accelerometer pitches and yaws during walk. Pitching of the accelerometer generates the mixture of the acceleration of gravity to the measured acceleration and causes error in the integrated data. The angle of the sensor system is measured by integrating the angular velocity derived from the gyro and is used to estimate the horizontal acceleration corresponding to the axis of the accelerometer. Although the yawing is caused by the installation of the sensor system and the rotation of the foot, the size of the horizontal acceleration projected to the horizontal plane is hardly affected. The integration of the horizontal acceleration is performed in every stance phase. The integral constant of the horizontal velocity must be zero during the stance phase and that of the angle of the sensor system is derived from the horizontal acceleration. Experimental results show that the difference between the real distance that subjects walked and the estimated one is 0.2[%] for seven subjects on average. The estimated distance is hardly affected by careless placement of the sensor system to the foot and abnormal walk such as fixed leg or shuffling.View full abstractDownload PDF (2604K) -
Kimio SASAKI, Yoshihiro SEITA2000 Volume 36 Issue 11 Pages 916-922
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper we psesent a new non-uniform sampling theorem for band-limited signals, together with an associated interpolation formula of non-uniformly sampled sequences, and propose a practical algorithm for reconstructing sampled sequences of any uniform interval from a non-uniform one by using the frequency sampling technique and FFT algorithm. After a brief review of the ordinary uniform sampling theorem, the non-uniform sampling theorem is given, together with an aliasing free condition being made clear. Based on the fundamentals, an interpolation formula of non-uniformly sampled sequences is derived and a practical implementation algorithm to reconstruct uniformly sampled sequences of any desired interval from a non-uniformly sampled version is proposed by way of the frequency sampling technique and FFT algorithm. The principle as well as the effectiveness of the proposed algorithm are ascertained through numerical experiments with sinusoidal and FM signals of finite bandwidths.View full abstractDownload PDF (1874K) -
Yasuhide KOBAYASHI, Tsuyoshi OKITA, Hiroaki MUKAIDANI2000 Volume 36 Issue 11 Pages 923-929
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA Wiener system, i.e., a system comprising a linear dynamic subsystem and a nonlinear memoryless subsystem connected in a cascade, is identified. The linear subsystem is expressed by the ARX model and the order is unknown. The class of all possible nonlinearity is so wide that it cannot be presented in a specific form. The artificial neural networks which have the ability to learn complex nonlinear relationships provide an ideal means of modelling complicated nonlinear systems. The artificial neural networks are applied to represent these nonlinear elements. The connection coefficients of the artificial neural networks and the parameters of the ARX model are related each other, then they can not be obtained independently. So, they are estimated simultaneously based on the maximum likelihood method. The order of the linear subsystem and the structures of the neural networks are selected by the minimum description length. It is demonstrated in digital simulation that the proposed identification technique is efficient in the system with high level nonlinearities.View full abstractDownload PDF (1498K) -
Takao HINAMOTO, Takashi KOUNO2000 Volume 36 Issue 11 Pages 930-936
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn the implementation of a state-estimate feedback controller with the finite word length of fixed-point arithmetic, an L2-sensitivity measure is derived for evaluating the overall sensitivity of a closed-loop transfer function with respect to the state-estimate feedback controller's coefficients by using a pure L2-norm. An iterative procedure is then presented for synthesizing the state-estimate feedback controller's structures that minimize the L2-sensitivity measure. The present L2-sensitivity minimization is more natural and resonable than the existing L1/L2-mixed sensitivity minimization. Furthermore, the above iterative procedure has the advantage of analytically calculating the estimate at each iteration. Finally, a numerical example is given to illustrate the utility of the proposed technique.View full abstractDownload PDF (1305K) -
Guisheng ZHAI, Bo HU, Kazunori YASUDA, Anthony N. MICHEL2000 Volume 36 Issue 11 Pages 937-942
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, we investigate the L2 gain properties of time-controlled switched systems consisting of several linear time-invariant subsystems by using an average dwell time approach incorporated with a piecewise Lyapunov function. We show that when all subsystems are Hurwitz stable and have the L2 gain smaller than a positive scalar, the switched system can achieve any L2 gain larger than such scalar if the average dwell time is chosen sufficiently large. The result is extended to the case where state jumps occur at switching points, and to a class of nonlinear switched systems by considering a Hamilton-Jacobi inequality for each subsystem.View full abstractDownload PDF (1300K) -
Iterative AlgorithmsTakashi SHIMOMURA, Takayuki TAMAKOSHI, Takao FUJII2000 Volume 36 Issue 11 Pages 943-951
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn order to reduce the conservatism of LMI-based multiobjective control design, we attempt to solve mixed problems with non-common LMI solutions, while focusing on the state-feedback case, in particular, H2/H∞control, regional pole placement, and their joint problem. Completing the square in terms of the controller variable and replacements of negative semi-definite quadratic terms by their upper bounds provide a new technique to convert BMIs into LMIs. A suitable choice of the parameters in these upper bounds with iterative algorithms guarantees their convergence property. Through an illustrated example, H2 performance is found to be considerably improved.View full abstractDownload PDF (1606K) -
Taek-Kun NAM, Tsutomu MITA2000 Volume 36 Issue 11 Pages 952-961
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, the multirate digital control will be applied to the control of nonholonomic chained systems with multiple chains. Then a new method called a variable rate deadbeat control is proposed to control high order chained systems. This control has a potential to avoid an exessive overshoot and prevent the trajectories to pass close to the singular points of the transformation needed to obtain the chained forms. These control strategies will be applied to the posture control of an under water vehicle and underactuated manipulator.View full abstractDownload PDF (1597K) -
Masashi NAKAMOTO, Takashi KOKUBO, Akinori KAMITO, Keiko SHIMIZU2000 Volume 36 Issue 11 Pages 962-971
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA control system for a nitrogen oxide (NOx) decomposition process in a combined cycle power plant is developed using a Generalized Predictive Control (GPC) method and a Linear Quadratric Regulator (LQR) method. NOx in gas turbine exhaust gas flow is decomposed by ammonia (NH3) at catalysts. NH3 flow rate is adjusted to keep the NOx flow rate at a setpoint. The control system has a cascade scheme that includes NOx flow control designed by a GPC method and NH3 contorol designed by an LQR method. Experimental results on an actual plant during commercial operations show not only control performances but also practicability.View full abstractDownload PDF (2177K) -
Hiroshi SHIMODA, Yoshiaki NIKAIDO, Naoki UMEDA, Naoki HAYASHI, Hidekaz ...2000 Volume 36 Issue 11 Pages 972-979
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSHead-attached interface device (HIDE) has been developed as a new type of wearable computer in an actual work environment. HIDE has flexible integrated functions which consist of speech recognition and view direction detection as hands-free operation methods, see-through display and audio guidance as information presentation methods. In this paper, the design concept of HIDE is first described, and the individual functions to configure the design concept and overall configuration are explained. Then, an integrated prototype system of HIDE has been produced, and its functional evaluation tests were conducted to examine the individual functions with respect to resolution of displayed image, speech recognition rate, and detection rate/time of view direction. Finaly, a WWW browser was made as an example application software for testing and repairing electric circuit board, in order to confirm the integrated functions of HIDE.View full abstractDownload PDF (3887K) -
Joo Kooi TAN, Seiji ISHIKAWA, Kiyoshi KATO2000 Volume 36 Issue 11 Pages 980-984
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA motion capturing technique based on factorization is presented in this paper. Factorization is a novel technique for recovering 3-D shape of a rigid object from a single video image stream. The technique is extended in this paper to motion capture by employing multiple settled video cameras and by defining an extended measurement matrix which describes motion of an object. The main advantage of the proposed technique is that one does not have to calibrate the video cameras, since the camera parameters are not used in the recovery calculation. Another advantage is that it recovers the motion of an object during observation all at once and therefore alignment of the motion in the time axis is not necessary. The technique is applicable to any object irrespective of rigid or non-rigid. In the performed experiments, several human motions have been successfully recovered in a 3-D way by the proposed technique.View full abstractDownload PDF (2903K) -
Masashi OHATA, Tsuyoshi TOKUNARI, Kiyotoshi MATSUOKA2000 Volume 36 Issue 11 Pages 985-994
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper proposes a new online algorithm for blind separation of mixed signals. The algorithm doses not any knowledge on statistical properties of the source signals. It estimates the distribution of sourses based on Gaussian mixture model and separates the observed signals. Although similer algorithms can be seen in literature, they are batch type and are not applicable for nonstationary sources. Since the proposed algorithm is an online algorithm, it is applicable for nonstationary sources as well as stationary ones.View full abstractDownload PDF (1924K) -
Toshiro KUMON, Tatsuya SUZUKI, Makoto IWASAKI, Tomonori HASIYAMA, Nobu ...2000 Volume 36 Issue 11 Pages 995-1002
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn case that the information of the structure of the system (i.e. order of the system) is not available, a designer must find a suitable structure to represent the system behavior appropriately. Generally speaking, this problem results in a large scale combinatorial search problem. One of the efficient way to solve this kind of problem is to utilize a so-called evolutionary computation. In this paper, we propose a new modeling method for structureunknown multi-input single-output linear dynamical systems using evolutionary computation. However, even if we use the evolutionary computation, an unreasonably long computational time is often required because of its vast search space. In order to overcome this problem, we try to set the limit of the search space based not on the knowledge of the plant but on AIC (Akaike's Information Criterion). After this process, we apply a genetic algorithm to find a sub-optimal model structure. We also carry out some numerical experiments to show the usefulness of our proposed method from viewpoint of both validity of the structure and accuracy.View full abstractDownload PDF (2523K) -
Yasue MITSUKURA, Minoru FUKUMI, Norio AKAMATSU2000 Volume 36 Issue 11 Pages 1003-1008
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, we propose a method to design of a neural network (NN) by using a genetic algorithm (GA) and simulated annealing (SA). And also, in order to demonstrate the effectiveness of the proposed scheme, we apply the proposed scheme to a coin recognition example.
In general, as a problem becomes complex and large-scale, the number of operations increase and hardware implementation to real systems (coin recognition machines) using NNs becomes difficult. Therefore, we propose the method which makes a small-sized NN system to achieve a cost reduction and to simplify hardware implementation to the real machines.
The coin images used in this paper were taken by a relatively cheap scanner (scanning 12mm in width). Then they are not complete, but a part of the coin image could be used in computer simulations. This is the reason why the width of coin images is limited. If the conventional scheme was used for this image, it would have low recognition accuracy. Therefore, in order to obtain high recognition accuracy, we propose a new scheme.
Input signals, which are Fourier spectra, are learned by a three-layered NN. The inputs to NN are selected by using GA with SA to make a small-sized NN. Simulation results show that the proposed scheme is effective to find a small number of input signals for coin recognition.View full abstractDownload PDF (2757K) -
Takao KUDOH, Takahiro HANYU, Michitaka KAMEYAMA2000 Volume 36 Issue 11 Pages 1009-1018
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSNew logic-in-memory architecture of a high-performance road-extraction VLSI processor is proposed to solve data transfer bottleneck between a memory and processing elements. A VLSI-oriented algorithm having parallelism and regularity is developed to check whether a car can go through the 3D topographical map obtained from 3D instrumentation. To execute the algorithm very fast based on locally parallel processing, a highly parallel VLSI processor with many redundant processing elements is discussed. The processing element consists of an arithmetic element and a register for the storage of the topographical data. A shift register is constructed in the processing element array, so that data transfer between adjacent registers is effectively done. The topographical data once read out from a memory is being stored in shift registers and it is reused until the related processing is completed, so that memory bandwidth becomes minimum. Allocation such that data transfer bottleneck between processing elements becomes minimum is discussed, and it is made clear that the logic-in-memory architecture based on dynamic active control of the redundant processing elements is very useful for the high-performance parallel processing. Finally, the evaluation of the proposed VLSI processor is done, and its superiority to other equivalent processors is made clear from the viewpoint of performance and hardware complexity.View full abstractDownload PDF (5051K) -
Makoto YASUDA, Takeshi FURUHASHI, Tomonori HASHIYAMA, Shigeru OKUMA2000 Volume 36 Issue 11 Pages 1019-1028
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSOne of the major problems of the technique that controls actions of agents by potential functions is to determine the function shapes suitable for the environment. The authors have proposed a method for autonomous acquisition of the parameters of potential functions by the agents with reinforcement learning, especially classifier systems, in multi-agent environment.
In this paper, we study a method to introduce temporary potential functions, which agents can set at desirable positions, so as to improve controllability of other agents. Experiments with a pursuit game are done, and the following are confirmed: acquisition of the temporary potential parameters for capturing the prey agent is possible; acquisition of the temporary potential parameters for the hunter agents to control each other and cooperatively capture the prey agent is also possible; the temporary potential increases the capturing probability of the prey agent.View full abstractDownload PDF (3499K) -
Yuichi KOBAYASHI, Jun OTA, Kousuke INOUE, Tamio ARAI2000 Volume 36 Issue 11 Pages 1029-1036
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSTo apply reinforcement learnig in the real world, we need to process sensor data adequately for action learning. Since it is difficult to construct state space and to learn the appropreate action simultaneously, we assume that an evaluation is given to each step of action. Evaluations are binary signals that mean actions are good or bad. Under this condition, we propose a method of dividing and clustering the state space. The TRN (Topology Representing Networks) algorithm is a vector quantization algorithm, and it can preserve topology in the input space. We apply the TRN algorithm to our problem with dynamically increasing nodes and the radial basis function.View full abstractDownload PDF (3480K) -
Masami IWATSUKI, Daiji KITAGAWA, Yusuke KAWANO2000 Volume 36 Issue 11 Pages 1037-1043
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper proposes a new robust method of stereo matching using coarse to fine hierarchical color segmentation of each scanline in stereo images. The proposed method can unambiguously decide disparity even though repetitive patterns exist in the images, because the longest sequences with same color pattern in both scanlines are simply regarded as matching regions. Furthermore, our matching algorithm does not need the nonreversal constraint in correspondence. Therefore, this method is effective for the scene including narrow occlusion.View full abstractDownload PDF (5068K) -
Toshio ENDO, Hiroshi WATANABE, Tomikazu YAGI, Masashi MIZUNO2000 Volume 36 Issue 11 Pages 1044-1046
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe induction heating inspection is versatile to detect surface defects on square steel billets. Their bright scratches have poorer emissivities than as-rolled skins, which interferes the temperature mapping to spot defects. Authors sprayed powders elecrostatically before heating. The powder coat effectively levels the emissivity all over and has enhanced the accuracy in an on-line inspection.View full abstractDownload PDF (2014K) -
Toshihiko KOYAMA, Hiroyuki INABA, Yasumasa ITAKURA, Masao KASAHARA2000 Volume 36 Issue 11 Pages 1047-1049
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSWe report the experimental results of debris flow detection based on MPEG video. A detection algorithm utilizes MPEG motion vectors. The advantages of the system are confirmed by simulation with some real debris flow movies.View full abstractDownload PDF (1094K) -
Guisheng ZHAI, Shigemasa TAKAI, Kazunori YASUDA2000 Volume 36 Issue 11 Pages 1050-1052
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, we consider a controller failure time analysis problem for linear time-invariant systems. By using a piecewise Lyapunov function, we show that if the unavailability rate of the controller is smaller than a specified constant and the average time interval between controller failures is large enough, then global exponential stability of the system is guaranteed.View full abstractDownload PDF (480K) -
Jiankun WANG, Zenta IWAI, Mingcong DENG2000 Volume 36 Issue 11 Pages 1053-1055
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper is concerned with the design method of a decentralized vibration control system for multi degree of freedom structures. The method is based on the partial model matching on frequency domain by minimizing the relative model error functions between the diagonal elements of the open loop transfer function matrix of the control system and those of the reference model. The effectiveness is examined by applying the proposed method to active vibration control of a 4th degree of freedom structure.View full abstractDownload PDF (469K) -
Reiko TANAKA2000 Volume 36 Issue 11 Pages 1056-1058
Published: November 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper deals with symmetric control systems consisting of some modules with local internal symmetries. The symmetry of the whole system can be described with a direct product G×L of the groups G for global symmetry and L for local symmetry. As a characteristic determined from the two symmetries, we focus on the loss of controllability in the systems by certain symmetric failures and clarify that the conditions for the loss of controllability are given for each level of symmetry separately.View full abstractDownload PDF (484K)
- |<
- <
- 1
- >
- >|