IEEJ Transactions on Electronics, Information and Systems
Online ISSN : 1348-8155
Print ISSN : 0385-4221
ISSN-L : 0385-4221
Volume 120 , Issue 4
Showing 1-18 articles out of 18 articles from the selected issue
  • Kenji Shirai, Yoshinori Amano, Kazuo Inoue
    2000 Volume 120 Issue 4 Pages 453-463
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    The paper addresses the problem of routing control of transaction in a parallel queuing system with transaction lost where customers must begin service within given deadlines. For a system of N parallel servers with probabilistic routing scheme, the optimal routing depends upon only the probabilistic characteristics in transaction lost dynamics, under the transaction loses in each route are modeled by Ito's stochastic differential equation.
    Download PDF (1025K)
  • Kenichiro Hayashi, Akifumi Otsubo, Kazuhiko Shiranita
    2000 Volume 120 Issue 4 Pages 464-471
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    In this paper, we propose a tuning method in which the membership functions in a fuzzy controller can be set up simply by utilizing the knowledge of tuning parameters in a conventional PID controller. In the method proposed here, fuzzy control is realized by applying the simplified inference method to only four simple fuzzy control rules, in which the piecewise linear membership functions determined by the knowledge are used in the antecedent part.
    Therefore, we first show that PI control can be realized by using the linear membership functions in the antecedent part of the above fuzzy control rules. Then, we show that the piecewise linear membership function used in the proposed fuzzy control can be determined easily by the composition of two different linear membership functions. Finally, from the results of the simulation for the first- and second-order lag systems with dead time, we show the effectiveness of this fuzzy controlin improving control performance, compared with a conventional type of fuzzy control using the triangular membership functions.
    Download PDF (1724K)
  • Kazuhiko Shiranita, Kenichiro Hayashi, Akifumi Otsubo
    2000 Volume 120 Issue 4 Pages 472-480
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    This paper describes a system for automatically sketching patterns on the surface of porcelain. For the implementation of sketching on porcelain, we propose a sketching method which uses the numerical control data output from the CAD system. This system comprises two parts. One is a measurement system for the three-dimensional shape of the porcelain piece using the Slit-Light Method, and the other is an automatic pattern sketching system. Since porcelain pieces are classified into two types according to their shape, two three-dimensional measurement systems are developed. For these measurement systems, we propose a method for calculating the values of three-dimensional coordinates in a simple and fast manner. In the sketching system, the sketching is performed with numerical control. The data for the numerical control used for sketching patterns onto the porcelain are generated using the results measured by the measurement system and CAD system.Experimental results of sketching on porcelain using the numerical control data show that the system is effective.
    Download PDF (5042K)
  • Hong Zhang, Masumi Ishikawa
    2000 Volume 120 Issue 4 Pages 481-488
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    In this paper, we propose a novel approach to dynamic inverse optimization problems by the learning of neural networks. Dynamic inverse optimization problems here mean to estimate a criterion function under which given sequences of input and output of a dynamical system are optimal. A neural network architecture representing the optimality condition including an algebraic Riccati equation is proposed for solving dynamic inverse optimization problems. Applications of the proposed method to observed input and output sequences well demonstrate its effectiveness.
    Download PDF (1093K)
  • Yasuhiro Kobayashi, Toshiyuki Sawa, Toshiyuki Furukawa, Toshiyuki Fuji ...
    2000 Volume 120 Issue 4 Pages 489-500
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    An improved search method has been developed by incorporating an agenda (solution pool) mechanism into the tabu search framework and it was evaluated through application to a job shop scheduling problem. The mechanism facilitates search from a solution in the agenda, while neighbor solutions being sequentially searched by the conventional tabu search. The tabu-list is dynamically synthesized for a selected solution, since the individual history of the solution is quite different from the actual search process.
    The bench mark problem was the shortest completion of 6 jobs with 6 machines. Case studies with 50 runs were done for the developed method and conventional search methods. The results reveal that optimal solutions are more frequently (62%) obtained by the developed method than by a conventional tabu search (36%) and a genetic algorithm (46%).
    Download PDF (3847K)
  • Teppei Komori, Yukihiro Fujiwara, Masayuki Fujita, Kenko Uchida
    2000 Volume 120 Issue 4 Pages 501-506
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    We propose an approach to Automated Driving based on Visual Servoing. Visual servoing is a vision-based control architecture for target tracking without recovering 3D scene in workingspace. To realize the automated driving, we design two controllers: one (lateral controller) is for controlling lateral vehicle dynamics and the other (longitudinal controller) is for controlling longitudinal vehicle dynamics. We first derive both models of lateral vehicle dynamics and camera, and design a lateral controller based on visual servoing. Next, we design a longitudinal controller without a model of longitudinal vehicle dynamics. Finally, we show efficiences of our automated driving by conputer simulations.
    Download PDF (1207K)
  • Hiroki Yoshimura, Tadaaki Shimizu, Naoki Isu, Kazuhiro Sugata
    2000 Volume 120 Issue 4 Pages 507-515
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    An adaptive noise reduction filter composed of cascaded sandglass-type neural networks (CSNNRF) is proposed. A given number of unit sandglass-type neural networks (SNN), each of which has a three-layer structure and consists of a same number of neural units in the input and the output layers and a single neural unit in the hidden layer, are connected in cascade. The number of unit SNNsis adaptively determined so as to be equal to a rank of covariance matrix of an original noise-free signal (signal component). Outputs of hidden layer units in individual unit SNNs, whose variances are equivalent to eigenvalues of the covariance matrix of the observed signal, are statistically compared by use of ANOVA (analysis of variance) to estimate the rank. When a CSNNRF is composed of the number of unit SNNs equal to the rank, the observed signal is filtered without any loss of signal component while noise component is maximally reduced. It was shown by computer experiments that the rank was almost always estimated accurately in an adaptive manner, and that noise reduction from the signal was carried out optimally.
    Download PDF (1774K)
  • Soichiro Hayakawa, Nuio Tsuchida, Hitoshi Tanaka
    2000 Volume 120 Issue 4 Pages 516-521
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    Shaft inserting is frequent and important operation for the automatic parts assemble. To insert a shaft, the active method using the force sensor and the passive method using RCC (Remote Center Compliance) system have been studied. We develop the sensor using the cross type PSD as the position sensor and develop the shaft inserting system for the moving object using the robot with this new sensor. By using the robot with this sensor, the robot can insert a shaft into a hole of the moving object with 100μm of the clearance at a speed of 40mm/s. Our proposed system is effective for the shaft inserting into the hole of the moving object.
    Download PDF (3470K)
  • Akio Uekawa, Yukio Goto, Masaharu Umezu, Haruki Furusawa, Ichiro Nakah ...
    2000 Volume 120 Issue 4 Pages 522-528
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    It takes a lot of CPU time to calculate the optimal routes between the start links and the destination links in a large scale road network.
    We have developed a fast method for route calculation using pre-determined routes between areas. In this method, a large scale road network is divided into many areas. The main parts of the optimal routes from each start area to all destination areas are pre-determined and stored in the data base (area route data base). The route calculator proposed in this method calculates the optimal route from a start link to a destination link using the pre-determined routes from the corresponding start area stored in the area route data base.
    This paper describes a way of calculating the exact route using area route data base. Experimental results compared with simple use of Dijkstra's method show that the proposed methodis quite effective for reducing the route calculation time.
    Download PDF (3738K)
  • Minoru Fukumi, Norio Akamatsu
    2000 Volume 120 Issue 4 Pages 529-535
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    This paper presents a method to extract rules from multilayered neural networks (NNs) formed using evolutionary algorithms (EAs). The objective of this study is to extract rules from NN, achieving 100% recognition accuracy in pattern recognition systems. NNs to be extracted rules are formed using EAs. Hy-brid algorithms of NN and EAs perform the formation of small-sized NN systems, which are suitable for rule extraction. EAs in this paper are a random optimization (search) method (ROM) and a genetic algo-rithm (GA). In this paper iris data and coin data sets are used as inputs. EAs are utilized to reduce the number of connection weights in NNs. The network weights survived after the training by ROM and GA represent regularities to perform pattern classification. The rules are then extracted from the networks in which hidden units use signum functions to produce binary hidden outputs, while sigmoid units are used in NN training. It enables us to extract simple logical functions from the networks. It is shown by means of computer simulations that this approach is more effective in rule extraction than conventional methods.
    Download PDF (1512K)
  • Hirokazu Okano, Hideki Imai, Hiroyuki Une, Yasuaki Tannaka
    2000 Volume 120 Issue 4 Pages 536-545
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    Nowadays, Internet/Intranet, WWW and E-mail are widely used. Therefore usable security systems based on encryption are desired.
    In this paper, we propose multimedia information partial encryption systems which enable free selection of confidential information of processing data on a screen, and encrypting them. Therefore, we can construct encryption systems which are superior in view of both confidence and operation. For example, in a text, only part of confidential information is encrypted and further translated into text mode which is ASCII code, and then returned into the text. Therefore, the text including encrypted data can be handled easily as a text file by ordinary word processors. Also, in databases, users can execute encrypting/decrypting data of each items when data are entered. These systems can also be used for images, spreadsheets, and digital signatures. Each part of confidential information has its own security level and a unique encrypting key. In these systems, security levels adapted to business organizations can be allocated, for example, a user in charge or the company president can have the highest security level.
    Download PDF (5922K)
  • Sang-Gu Lee, Jong-Gyu Han, Hee-Hyol Lee, Michio Miyazaki, Kageo Akizuk ...
    2000 Volume 120 Issue 4 Pages 546-553
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    In this paper, we implement a land cover pattern classification system for remotely sensed data by using a neuro-fuzzy algorithm, and compare it with the conventional methods of the Back-Propagation learning and the Maximum-Likelihood algorithm. The neuro-fuzzy pattern classifier has a 3-layer feed-forward architecture that is derived from a generic fuzzy perceptron. The digital image used in our research was acquired with the AMS (Airborne Multispectral Scanner). We determine the eight classes covered the majority of land cover feature on Daeduk Science Town. The results show that the proposed classifier is considerably more accurate to the mixed composition area with complex classes.
    Download PDF (3703K)
  • Kenji Shirai, Yoshinori Amano, Kazuo Inoue
    2000 Volume 120 Issue 4 Pages 554-560
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    The purpose of this paper is to theoretically identify the condition of optimal control problem for the Queuing system with transaction lost in the stationary. In order to prove this, we formulated and analyzed the optimal condition as the Financial Cost Function. Accordingly, we make clear this optimal condition for the system. We introduce the concept of Finance Theory in order to formulate the Financial Cost Function.
    Download PDF (1018K)
  • Ping Wang, Takahiro Yamasaki, Masaki Yumoto, Takenao Ohkawa, Norihisa ...
    2000 Volume 120 Issue 4 Pages 561-565
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    A stochastic qualitative reasoning method has been developed for fault diagnosis of building air condition-ing systems. In this method, the fault part can be identified by ‘agreement rate’, which is a parameter that shows how reasoning behavior reflects observations. However, when agreement rate is calculated at a certain point of time, reasoning must be done with several unit times from that point. Accordingly, it is inefficient to do reasoning several times at each point.
    In this paper, a continuous qualitative reasoning method is proposed. In this method, reasoning is done from the beginning of observations to the end. In addition, a technique for estimating the minimum exis-tence probability, which is used to remove the states with small occurrence, is introduced. As results, the performance has increased about 40 times in comparison with previous method.
    Download PDF (998K)
  • Seiichi SAITO, Shuichi NITTA
    2000 Volume 120 Issue 4 Pages 566-574
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    A high-speed bus circuit with low-power-consumption has been required among computer systems, especially in mobile use. This paper proposes an efficient and effective method to improve the analysis of bus circuits to be adapted to a higher-speed transmission with low-power-consumption. The proposed method is made by analyzing pulse waveform distortion focusing the relationship between the harmonics of the fundamental frequency of the pulses and the frequency response of transmission line. Further, in this paper, the looped bus circuit with non-power-consuming termination is proposed, which can improve the waveform distortion dramatically and realize both high-speed and low-power-consumption even at the system that the location of a driver is changed among IC's connected on the bus line. The looped bus circuit with non-power-consuming termination proposed by authors can achieve a data transmission rate 400M transfer/s, whereas conventional bus circuits are below 100_??_200M transfer/s, verified by the above proposed method, SPICE simulation and experiment.
    Download PDF (1419K)
  • Erwin Avianto, Masayasu Ito
    2000 Volume 120 Issue 4 Pages 575-582
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    To develop a boundary extraction method, and to evaluate an appropriateness of the extracted boundary from ultrasound images are significant problems, because of the complex ultrasound characteristics and the obscurity of tissue boundary image. This paper presents a new method for an automatic tissue extraction from medical ultrasound image, whose extracted edge is optimal from the view point of histogram. We pro-pose a novel method of the automatic tissue boundary extraction by using fuzzy morphology in order to cope with the ambiguous and obscure ultrasound image. Regarding a quantitative evaluation which is considered as the statistical characteristic and the gradient or distribution of level from the original image, proper edges or boundaries are detected using multi structuring elements here. We clarify a relation of boundary with structuring element in connection to fuzzy morphology that is used in the boundary extraction.
    Download PDF (6293K)
  • Kohji Sasabe, Kazuhisa Yoshida, Osamu Fujiwara
    2000 Volume 120 Issue 4 Pages 583-589
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    The conversion from common to differential-mode noise was investigated. Two simple parallel traces with injecting the common-mode voltage were considered for the mode conversion. The effect of the mutual cou-pling between traces on the noise mode conversion was examined experimentally. Measurements suggest that the induced differential-mode voltage should change the logic level of sensitive ICs and also that the tight coupling among the traces gives the lower mode conversion. In order to investigate the findings theoretically, an equation for determining the noise mode conversion was derived. As a result, it was confirmed that the calculated differential-mode voltages agree with the measured ones.
    Download PDF (1142K)
  • Tagawa Kiyoharu, Fumiya Ohara, Haneda Hiromasa
    2000 Volume 120 Issue 4 Pages 590-596
    Published: April 01, 2000
    Released: December 19, 2008
    JOURNALS FREE ACCESS
    A method based on the interval analysis is given for finding every solution of the inverse kinematics problem of general manipulators. The method is an interval version of binary search and rather simple in principle; however, it has been slow for practical applications. Therefore, a main contribution of this paper is to speed up the primitive version of the method by introducing some skillful techniques. In an improved version named the Fast Interval Bisection (FIB) method, not only the number of operations but also the storage space of variables are reduced successfully according to the topological analysis of computational graph. Furthermore, in order to generate a complete program of the FIB from an arbitrary kinematics equation automatically, a kind of compiler system is also developed. The experimental evaluation of the FIB is performed on the inverse kinematics problem for a six-revolute-joint manipulator with 16 solutions.
    Download PDF (1319K)
feedback
Top