IEEJ Transactions on Industry Applications
Online ISSN : 1348-8163
Print ISSN : 0913-6339
ISSN-L : 0913-6339
Volume 118, Issue 3
Displaying 1-24 of 24 articles from this issue
  • Yasuo Morooka
    1998 Volume 118 Issue 3 Pages 275
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Download PDF (111K)
  • Harutoshi Ogai
    1998 Volume 118 Issue 3 Pages 276-279
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Download PDF (1364K)
  • Takeshi Furuhashi
    1998 Volume 118 Issue 3 Pages 280-283
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Download PDF (1508K)
  • Fumiaki Takeda, Toshihiro Nishikage
    1998 Volume 118 Issue 3 Pages 284-290
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Up to now, our group have proposed a new type of recognition method using neural network (NN) which is endowed with generalization and flexibility. Especially, this method is constructed by mask processing, mask optimization by genetic algorithm, and neuro-recognition board. In this paper, we discuss introduction of the NN to partial sensed pictures of the paper currency. However, the partial sensed pictures are different each other even though they are same category of the pattern recognition. When we recognize those pictures using a conventional single NN, it is supposed that the recognition ability is not enough in the market. Then we propose a new recognition system which is constructed by plural NNs linked as a cascade for the partial sensed pictures. First, we show a problem of recognition using the conventional single NN by the simulation results. Then, in order to solve the problem, we show a construction of the proposed recognition system. Finally, we discuss ability of the proposed system by the simulation results.
    Download PDF (2473K)
  • Kotaro HIRASAWA, Masaaki HARADA, Masanao OHBAYASHI, Juuichi MURATA, Ji ...
    1998 Volume 118 Issue 3 Pages 291-299
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Universal Learning Network(ULN) which is a generalized Neural Network, can be used to model and control large scale complicated systems. But ULN can not be applied to discrete event systems. Therefore a discrete event oriented learning network which is called Automaton Learning Network(ALN) has been already proposed. ALN has the same structure as ULN has. In this paper. a generalized type of ALN named Probability Automaton Learning Network(PrALN) and Possibility Automaton Learning Network(PoALN) are presented in order to realize an ALN with non-deterministic nature. In the simulations with a relatively simple model, we studied the fundamental characteristics of PrALN and PoALN.
    Download PDF (1342K)
  • Masanao Ohbayashi, Kotaro Hirasawa
    1998 Volume 118 Issue 3 Pages 300-307
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    As the nonlinearity and complexity of a nonlinear system increase, it may be more difficult to construct a controller by the mathematical control theory. In such cases, it is very effective to construct the controller by using Neural Network(NN), because NNs have capabilities of coping with the nonlinearity and complexity of the nonlinear systems. NN controllers are constructed through learning to minimize a criterion function under certain environments of the system. But NN contollers may not work well under a very different environment from the environment at learning stage. In other words, for example, NN controllers are usually made without considering the changes of the environment because NN controllers do not have a means to supress their influences. So, in case that the environment changes, NN controllers do not work well. In this paper a robust control system design method for the changes of the external inputs to the system is discussed using second order derivatives of Universal Learning Network.
    Download PDF (1584K)
  • Kazuhiko Takahashi, Takayuki Yamada
    1998 Volume 118 Issue 3 Pages 308-314
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    In this paper, we propose a heterogeneous hidden layer consisting of both sigmoid functions and RBFs (Radial Basis Functions) in multi-layered neural networks and present a method for implementing the neural network to control systems. Focusing on the orthogonal relationship between the sigmoid function and its derivative, a derived RBF that is a derivative of the sigmoid function is used as the RBF in the neural network, so the proposed neural network is called an ONN (Orthogonal Neural Network) and the function mapping with the ONN can be treated as a kind of an orthogonal function series model. Identification results using a nonlinear function confirm both the ONN's feasibility and characteristics by comparing with those obtained using a conventional neural network which has sigmoid function or RBF in hidden layer. Using the ONN, a parallel type neural controller, which uses both the ONN output and the conventional control output as an objective system control input, is proposed. Simulation results for discrete-time nonlinear SISO system demonstrate the applicability of the neural controller for controlling nonlinear systems and experimental results for controlling angular velocity of a DC servo motor demonstrate its usefulness for controlling practical systems.
    Download PDF (2430K)
  • Toshiya Kaihara, Susumu Fujii
    1998 Volume 118 Issue 3 Pages 315-321
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    In highly integrated manufacturing lines, intelligent transportation system plays an important role in effective manufacturing operation. Fully de-centralized transportation architecture with flexibility and robustness against environmental changes is required in the manufacturing line. This paper studies the abilities of multi-agent paradigm with simple architecture and primitive algorithm to move appropriately in relation to several process machines. Efficient individual behaviour is obtained in simulation model and self-organized work flow is formulated with coordinated motions between transport agents. An evolutional approach based on genetic algorithm increases the efficiency of the transportational behaviour with robustness. The transport agents are capable of robustness and reliability coping with the several manufacturing demands.
    Download PDF (1624K)
  • Kenji Terada, Jun'ichi Yamaguchi
    1998 Volume 118 Issue 3 Pages 322-328
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    To countt the passing people through the gate of the buildings or the entertainment halls is considered important for the office security or the marketing researches. In this paper, an automatic counting method of the passing people by utilizing color information is proposed. First, in the proposed method, the color images of the passing people are obtained in series. As a next step, pixels on an angled measurement line are picked up and arranged along the time axis. As a result, a space-time image which includes just the data of the passing people over the angled measurement line is generated. On this space-time image, the the distribution shape of the incoming persons is different from that of the outgoing persons. By detecting the distribution shape of the people data, the moving direction of the passing people can be recognized. Finally, each number of the incoming persons and the outgoing persons could be measured by counting the people data with each direction information. In this paper, we describe the algorithm of counting the passing people using the two space-time images and show some experimental results obtained by using a simple experimental system to verify effectiveness of the proposed method.
    Download PDF (4406K)
  • Tadashi Iokibe, Yoshitsugu Yonezawa, Minako Taniguchi
    1998 Volume 118 Issue 3 Pages 329-334
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    For predicting the flow into a hydro-electric power station, complex natural phenomena have to be dealt with, so conventional mathematical models based on hydraulics may not produce satisfactory results. When a neural network is used, its construction cannot be easily determined, and so extra neural networks have to be provided separately in addition to the normal neural network, according to experts' opinions about the problem. To solve these problems, the authors took the standpoint that if the inflow rate time-series data for hydro-electric power stations exhibit deterministic chaos, the status in the near future can be predicted. So the authors have applied the local fuzzy reconstruction method as a deterministic nonlinear short-term prediction method to data for the flow of water into hydro-electric power stations. In this paper, typical outflow analysis method using conventional mathematical models are first described briefly. Next, the “Local Fuzzy Reconstruction Method” is described. Third, chaotic behavior of water flow data into hydro-electric power stations are illustrated. Finally, the results of applying the method to the prediction of the flow into hydro-electric power stations are presented.
    Download PDF (1588K)
  • Nobuhiro Kataoka, Hisao Koizumi, Hideaki Simizu
    1998 Volume 118 Issue 3 Pages 335-344
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Conventional host-based central-processing type logistics information systems collect all information about stocked products (sales results, inventory, out-of-stock items) on a single host computer, and based on this information perform ordering, shipping, receiving, and other processing. In a client/server architecture, the system is not simply downsized: in order to ensure more effective use of logistics information and closer coordination with manufacturing information systems, the logistics information system must be configured as a distributed system specific to a given factory and its various products. Such distributed systems each function acts independently, but at the same time the overall system of which they is part must operate in harmony to perform cost optimization, adjust allocation of resources among different factories and business locations, and present a single monolithic interface to retailers and sales agents. In this paper, we propose a logistics information system with a distributed architecture as well as agents whose role is to coordinate operation of the overall system, as one means of realizing this combination of component autonomy and overall system harmony. The methodology proposed here was applied to a proving system, and its effectiveness was verified.
    Download PDF (1998K)
  • Hiroki Ishikawa, Yoshihiro Murai
    1998 Volume 118 Issue 3 Pages 345-352
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Recently, the soft-switching technique is used for the power conversion systems. This prevents the switching losses and EMI (electro-magnetic interference). Therefore, the authors had proposed a new series resonant dc link PWM (pulse width modulation) inverter. The prototype inverter had some merits that had small dc inductance and the output capacitors with fixed pulse frequency. However, the waveform of the output current was clamped at around zero with open loop control, and the voltage of the resonant capacitor had to be inverted. Because of these problems, the control became difficult. This paper proposes a series resonant dc link PWM inverter for the solution of these problems. The result of improvement, the control becomes easier, and the efficiency becomes higher than the prototype inverter. In this paper, the system, the operation, the experimental result, and the design of the improved inverter are explained.
    Download PDF (2012K)
  • Nobuyoshi Mutoh, Kenkichi Kagomiya, Toshiaki Kurosawa, Masahiro Konya, ...
    1998 Volume 118 Issue 3 Pages 353-362
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Horizontal vibrations of elevator cars mainly occur because a car swings as roller guides installed at corners of a car frame move on a winding guide rail at high speeds. Rider comfort in high speed elevators is worsened by these vibrations. Conventional active dampers suppressing horizontal vibrations using ac servo motors make cars heavier so driving power becomes larger, and they are not easily applied to existing elevators. An active damping control method suited to super-high-speed elevators is described which can solve these problems. The method suppresses vibrations by generating only enough magnetic force needed to suppress them only when vibrations of the car frame are produced. The vibrations are detected using acceleration detectors and magnets installed on left and right sides of the car frame. A computer simulator was made to analyze phenomena of car vibrations and to verify effects of the proposed magnetic damping controller. It was found that the vibrations generated on the cabin floor were remarkably large when left and right sides at the upper and lower parts of the car frame were swung by sine waves with the same phase. The vibrations had two resonant modes. Results obtained with the computer simulator and a full scale running simulator showed that the acceleration on the cabin floor, even at the resonant frequencies, could be reduced by the magnetic damping control to around 0.1m/s2 which would provide a comfortable ride.
    Download PDF (3012K)
  • Takao Goto, Muneaki Ishida, Takamasa Hori
    1998 Volume 118 Issue 3 Pages 363-370
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    This paper proposes an active filter using a series-resonant high-frequency AC-linked DC-AC converter, which compensates the harmonic currents in the utility line generated from the electric and electronic apparatus of the home appliances and discusses the ripple characteristics of the filter. First, an equivalent circuit which describes the macroscopic behavior of this converter is derived. Then, a method for the real-time output current control of the converter to compensate the harmonic currents in the utility line is realized based on the equivalent circuit of converter. The stability of the current feedback loop is analysed by Nyquist locus and Bode diagram to evaluate the ripple characteristics of the active filter. The usefulness of the proposed control algorithm is confirmed by analytical and experimental results.
    Download PDF (1567K)
  • Toshikazu Onda, Makoto Niwakawa
    1998 Volume 118 Issue 3 Pages 371-376
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    We have focused to identify three isolated closed-curves on the object to decide its position and orientation, because most of our objects are rigid and have some visible holes or dimples as closed-curves on their surface. The system uses three-dimensional closed-curves matching technique with stereo ranging, estimates its appearance in the scene using model-based feature matching, and then pick it up. The system can recognize and pick up randomly stacked casting parts in a bin.
    Download PDF (5283K)
  • Shizumasa OKUDAIRA, Kouki MATSUSE
    1998 Volume 118 Issue 3 Pages 377-384
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    It is to be desired that the operating frequency of a resonant inverter for induction heating is adjustable when various parts or uneven parts are treated by heating. In order to respond the demand, we reported a novel adjustable frequency quasi-resonant inverter and clarified the basic characteristics. However, the output current contained many harmonics and there was the possibility that the heat-treated depth in the uneven surface of the work did not become uniform unless the operating condition was optimum. The heat-treated depth depends on not only output frequency of the inverter but supplied power into the work coil. If the frequency and the supplied power can be controlled independently, the degree of freedom of the heat treated depth control increases. Therefor, we propose a frequency and power control method using the reported inverter. In the present method, the inverter and the induction heater are considered as a series RLC resonant circuit which includes an equivalent loss resistor and an equivalent output resistor. The input DC voltage of the inverter is so manipulated that the output power generated in the output resistor may be equal to the desired value after the operating frequency was controlled to the target value.
    Download PDF (1428K)
  • Toshihisa Shimizu, Akio Toba, Gunji Kimura
    1998 Volume 118 Issue 3 Pages 385-392
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    In order to improve the control performance and to make main circuit components small, high frequency switching strategy is applied to PWM inverters. On the other hand, switching loss on the switching devices tends to increase as the switching frequency goes high. To prevent the increase of switching loss, many kinds of resonant type inverter topologies have been proposed. It is called that an auxiliary resonant commutated pole (ARCP) inverter is one of the most suitable topology for PWM control and low switching loss operation. On the other hand, it is reported that efficiency of the ARCP inverter does not so high compared with that of the conventional so-called hard switching inverters on the practical applications. These discussion will be concluded if the theoretical loss estimation of the ARCP inverter is reported.
    In this paper, a theoretical loss calculation method and a optimum circuit parameter condition for the ARCP PWM inverter is proposed. It is clarified that the efficiency on the optimum circuit parameter condition will rather be increased but it is difficult to use the parameter in the practical application. Then, the modified parameter that is applicable to the practical circuit is obtained. Unfortunately, even on the modified condition, the resultant efficiency does not extremely high compared with that of the conventional hard switching inverter. These results are coincide well with the experimental results on the bread board set-up and the up to present efficiency which are reported on the practical application. By the way, ARCP inverters are going to recognized as the suitable strategies for reducing EMI noise. Then, the proposed loss estimation method and the practical parameter calculation method will be useful for designing the ARCP inverter with low EMI and high efficiency.
    Download PDF (1891K)
  • Takeshi Sugimoto
    1998 Volume 118 Issue 3 Pages 393-401
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Regenerative vehicles are working even on less busy lines. On such lines, when the train density is so low that regeneration is liable to be ineffective and, notably where the gradient section is long, the brake performance will be poor and the brake shoe will be heavily abraded. Therefore the resistor is mounted on the vehicles and the regenerated energy fails to be fully consumed, the residue is consumed at the resister. It is desirable not to mount the resistor on the vehicle and avoid an increased vehicle weight. Then, to mount or not should be decided with clever judgment.
    Therefore, we studied the effective coefficient of regenerative energy which provides the criteria for judgment. We looked for an estimation formula of effective coefficient of regenerative energy, and showed that changing the number of trains, the coefficient will not change so much. We proposed a technique of estimating manually and by a simple program based on that formula and confirmed that the calculation can be precise enough to be practically employed.
    Download PDF (1210K)
  • Yoshihisa Sato, Makoto Kyouto, Junji Kondoh, Satoshi Hanai, Ryuichi Sh ...
    1998 Volume 118 Issue 3 Pages 402-409
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Higher Magnetic field is of great advantage to Superconducting Magnetic Energy Storage because it promises a compact device. However, the serious problem of huge electromagnetic forces caused by high magnetic fields and large coil current, still remains. Electromagnetic stresses, reaching as much as several hundred bars, thus become a major factor in determining whether a large sized strong magnetic field system could be established. In order to solve this problem, we propose the force-balanced coil (FBC) concept that reduces the huge centering electromagnetic forces by balancing it with the hoop forces caused by toroidal current. Then we designed and fabricated a small size superconducting FBC system to demonstrate the FBC concept. It was successfully excited up to near rated current. It is shown that the FBC system has a potential for simplifying the supporting structures and making the facility compact.
    Download PDF (2316K)
  • Fumio Matsumura, Toru Namerikawa, Noriyuki Murata
    1998 Volume 118 Issue 3 Pages 410-416
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    An active magnetic bearing (AMB) is a bearing to suspend a rotor by magnetic forces without any contact. Feedback control is indispensable for a magnetic bearing, because it is essentially an unstable system. To design a feedback control system, a linear mathematical model of the plant is convenient, however the force of electromagnet is highly nonlinear. Then usually a linear approximation around the operating point is employed, but the obtained linearized model can not express the exact behavior of the system at any other operating points. In this paper, we discuss the wide area stabilization of magnetic bearing using exact linearization approach.
    At first, we derive a nonlinear mathematical model for a magnetic bearing, then we show this nonlinear system is belong to a class of exactly linearizable nonlinear systems. Next, by using exact linearization method, we transfer the nonlinear model of a magnetic bearing to a linear time invariant state model in spite of change of the operating point and the rotational speed of the rotor. Then we construct the state feedback control system by conventional LQ method. Finally, we evaluate the validity of our proposed method by experiment.
    Download PDF (4275K)
  • Keiji Wada, Hirofumi Akagi, Hideaki Fujita, Satoru Furutani
    1998 Volume 118 Issue 3 Pages 417-418
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Download PDF (899K)
  • Tadashi Fukami
    1998 Volume 118 Issue 3 Pages 419-420
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Download PDF (1027K)
  • Shinji Shinnaka
    1998 Volume 118 Issue 3 Pages 421-422
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Download PDF (172K)
  • Mikio Yasubayashi, Masami Ito
    1998 Volume 118 Issue 3 Pages 423-424
    Published: March 01, 1998
    Released on J-STAGE: December 19, 2008
    JOURNAL FREE ACCESS
    Download PDF (1089K)
feedback
Top