Transactions of the Society of Instrument and Control Engineers
Online ISSN : 1883-8189
Print ISSN : 0453-4654
ISSN-L : 0453-4654
Volume 18, Issue 4
Displaying 1-18 of 18 articles from this issue
  • Nobuaki KOBAYASHI, Takayoshi NAKAMIZO
    1982 Volume 18 Issue 4 Pages 317-323
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    The problem is studied in this paper of pole placement with state feedback while preserving noninteraction
    A modified canonical decoupled form of the system with r-inputs and m-outputs is first introduced and then it is shown that the number of assignable poles can be calculated Under a mild condition, a necessary and sufficient condition for existing a stable decoupled system is presented in terms of system invariant zeroes. A simple method to examine the stability of a decoupled system is also shown.
    Download PDF (813K)
  • Shinji HARA
    1982 Volume 18 Issue 4 Pages 324-328
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    New concepts for an unknown input unobservable subspacs and an unknown input oservability for discrete-time linear multivariable systems are introduced. These new concepts, namely (r, s) unobservable subspace and (r, s; K) observability, are extensions of conventional one. Algorithms for obtaining the (r, s) unobservable subspace are presented using both matrix and vector-space (geometrical) operations. Necessary and sufficient conditions for (r, s; K) observability are also derived and the relation between this and the canonical form is clarified. It is also shown that there exists a dead-beat K-observer, of which the output coincides with the partial state of the system Kx (i) after some time, if and only if the system is (r, s: K) observable for some integers r snd s. An algorithm for constructing this observer is developed in a simple matrix form.
    Download PDF (643K)
  • Tsuyoshi OKADA, Masahiko KIHARA, Yukio MIZUKURA
    1982 Volume 18 Issue 4 Pages 329-334
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    In man-machine systems like piloted aircrafts, human operating characteristics have great significance. But these human operating characteristics vary with individual and environment. It is desirable in most cases that these variations are reduced.
    In this paper, we propose and examine experimentally a method to reduce these variations by utilizing the perfect model following (PMF) system with optimal feedback gain which has capability of sensitivity reduction. We call this method standardization. First, frequency responses of human operators were measured and a transfer function to fit the measured data was determind. Then the PMF system was synthesized for model and plant which have the measured human transfer functions. Next, replacing the plant with human operator, we measured frequency responses of the subject with PMF system. Frequency responses of the subject without PMF system were also measured for comparison. The results of these measurements show that the scatterings of human operating characteristics by individuals are significantly reduced in PMF system, which implies that the objective of the standardization is realized.
    Download PDF (768K)
  • Yoshifumi SUNAHARA, Shin-ichi AIHARA, Fumio KOJIMA, Naoto FUKUWA
    1982 Volume 18 Issue 4 Pages 335-342
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    The purpose of this paper is to solve the filtering problem for distributed parameter systems where system uncertainties exist on the system boundary. The mathematical model of the system is firstly described by a stochastic integral equation with values in Hilbert spaces.
    Secondly, the nonlinear filter dynamics is derived with the aid of martingale representation technique. In order to realize the filter dynamics, an approximated version by invoking Gaussian assumptions is discussed.
    Finally, for the purpose of supporting the theoretical development established here, digital simulation experiments are demonstrated.
    Download PDF (960K)
  • Zenta IWAI, Makoto SATO, Akira INOUE, Kazuo MANO
    1982 Volume 18 Issue 4 Pages 343-348
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    This paper is concerned with the design of an adaptive observer to estimate state variables and identify parameters of an unknown linear single-input single-output system with desired rate of exponential convergence. The observer structure is based on Kreisselmeier type representation of non-minimal realization, but in this study the adaptive law is obtained by a new scheme. The adaptive law is determined by equations forcing the errors into decreasing exponentially. That is, the adaptive observer is designed with the exponential rate of convergence in respect to parameter mismatches. The convergence rate is related on small number of parameters of the observer and the rate is specified arbitrarily. A sufficient condition for this property is obtained. An example is given to show the effectiveness of this design scheme.
    Download PDF (760K)
  • Kimio KANAI, Tadayasu DEGAWA, Shigeru UCHIKADO
    1982 Volume 18 Issue 4 Pages 349-356
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    In recent years, an adaptive observer which estimates the parameters and the states of an unknown plant has been actively examined. But most of studies are contributed only to a single-input single-output system. The extension of this procedure to the multivariable case seems to be more difficult because the canonical form for such systems has much more complicated structure.
    This paper deals with the problem of desining an adaptive observer for a multi-input multioutput linear time invariant discrete system where only the inputs and outputs can be measured. In order to estimate unknown parameters of a plant, a scalar output representation is defined and the resultant multi-inputs single-output system is considered. For the estimation of output equation, two different algorithms are obtained by using a recursive least square estimation method.
    The first approach makes use of estimated states for parameter estimations. In the second approach, the output equation is transformed into the linear combination of unknown parameters and known signales, and the signal is used for parameter estimation, and the iteration number of convergence is also determined. At the end of this paper, the results of digital simulation for a third-order system using two approaches are presented.
    Download PDF (741K)
  • Fujio OHKAWA, Yoo YONEZAWA
    1982 Volume 18 Issue 4 Pages 357-362
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    In spite of the large number of design methods for model reference adaptive control system (MRACS) have been proposed by the researchers of stability theorem, only a few studies of MRACS for multi-input multi-output (MIMO) plant with time-delay have been published to date.
    In this paper, a design method for discrete MRACS is proposed based on Lyapunov's direct method. The controlled plant is a MIMO linear plant with time-delays. It is shown that the proposed adaptive algorithms gurantee the convergency of plant's output to reference model's response by introducing an auxiliary input vector. For the adaptive algorithms proposed here, only available input-output date of the plant are used and no information on unknown plant parameter is required. Finally, the validity of the proposed adaptive algorithms is demonstrated by numerical examples.
    Download PDF (668K)
  • Multi-Vibrator
    Kiyoyuki TCHIZAWA
    1982 Volume 18 Issue 4 Pages 363-370
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    The purpose of this thesis is to deal with jump phenomena in nonlinear systems from a topological point of view. As a typical example of such a nonlinear system we take up an electric circuit system called multi vibrator.
    In the beginning, we explain ordinary phase plane analysis of the above system in order to compare with topological analysis and then describe jump phenomena on a phase plane. Secondarily, we represent multi vibrator system as an oriented graph with 6 nodes and 9 branches. The currents and the voltages in the branches of the system can be specified on 9-dimensional vector space respectively. Consequently we consider the system as two manifolds in R18. So we construct two differentiable manifolds which are 9-dimensional Kirchhoff space K9 and 9-dimensional branch characteristics space 9 where K9 is a linear subspace determined by the Kirchhoff laws and 9 is a space specified by resistor branches and capacitor branches. Then the dynamical system is described on the configuration space Σ9=K9∩ ∧9 which inverse function theorem guarantees. Moreover we derive the equation of the state on which two manifolds K9 and 9 are nontransversal systematically. As a result of this derivation we can see constructively that the transversality condition does not hold on the jump points of multi vibrator. It is also shown, however, that there is a state on which jump phenomena do not occur though the transversality is destroyed. This fact gives us a conjecture that Morse Index will classify the jump points. The above approach will be applied to not only electric circuit systems but also other types of nonlinear systems.
    Download PDF (871K)
  • Tatsuo NARIKIYO, Masami ITO, Shigeyuki HOSOE
    1982 Volume 18 Issue 4 Pages 371-376
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    In this paper dynamic systems with certain holonomic constraints are precisely modeled by introducing the penalty function and identifying it with potential energy. In this method the constraints are given as a manifold in a configuration space and actual motion is given as a smooth curve on the manifold.
    Application to 7-freedoms biped model shows that this method is sufficiently applicable to multi-freedom dynamic systems imposed certain nonlinear constraints. Still more, it is shown that numerical solutions obtained by the application of this method contain high frequency components. Therefore stability considerations of the numerical solutions and its error analysis become very important. But if step size of integration in Runge-Kutta method is suitably selected for some N (i.e. weight of penalty), errors of solutions can be reduced to small. And by precise analysis of linearized system it is shown that the high frequency parts of the solutions correspond to constrained variables whose amplitudes decrease in proportion to order (N-1/2) or more rapidly, as N→∞.
    Download PDF (754K)
  • Akira KOBAYASHI, Yasuhiko SAKAMOTO
    1982 Volume 18 Issue 4 Pages 377-384
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    This paper proposes a new method of constructing interval scales from numerical data obtained by paired comparisons. This method is based on band function model.
    Band function model has been suggested by Kobayashi and Shibata to construct a reasonable interval scale. However, Kobayashi-Shibata's iterative algorithm gives only one scale of various scales which should be given as solutions, and the accuracy is affected by the choice of initial value.
    First, the authors explain that information contained in paired comparisons data can be transformed reasonably into simultaneous homogeneous linear inequalities.
    Next, effective algorithm is introduced to solve the inequalities to obtain interval scales.
    The solution domain is given as a set of inner points of a polyhedral convex cone, and expressed with a set of edge vectors. As the extent of the solution domain indicates the fuzziness, further discussion on precision has become able to be done.
    This method is compared with Kobayashi-Shibata's method, and the solution of the latter method is shown to be located a step inside of the boundary of this method's solution domain.
    Finally, the proposed method is applied to the experimental data obtained from subjective comparisons of weights. The obtained interval scale shows good agreement with the real weighing scale.
    Download PDF (960K)
  • Shigeaki SAWADA
    1982 Volume 18 Issue 4 Pages 385-389
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    The triple point of indium was realized by sealed glass cells for the standard platinum resistance thermometry.
    Three samples of pure indium (about 700g, respectively) from different sources were put into cylindrical glass cells (external diameter; 35mm, height; 200mm, inner diameter of the thermometer well; 11mm), which were sealed off after evacuating lower than 0.01Pa. The cell surrounded by a hollow block of aluminum was set in an electric furnace. The melting and freezing curves were obtained by three long stem-type Pt resistance thermometers which had been calibrated at the zinc point, the tin point and the triple point of water.
    The estimated value of the triple point of indium from the melting and freezing plateaus gave 156.6296°C±0.0003°C on IPTS-68. The range of super-cooling in the freezing point experiment was as small as 1°C.
    Download PDF (775K)
  • The Flow-Coefficient of Area-Type Flowmeter
    Junji KANAZUMI
    1982 Volume 18 Issue 4 Pages 390-393
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    Characteristics of the area-type flowmeter have so far been investigated by many workers. However, our knowledge is at present far from complete owing to the difficult nature of the fluid flow involved. One of the problems yet to be studied may be the effect of the Reynolds number (combined effect of viscosity, density and speed of the flow) on the flow-rate coefficient of this type flowmeter. In the present paper is disscussed the structure of the internal flow at very low and very high Reynolds numbers, and theoretical formulae for the flow-rate are derived for respective cases. Two formulae are then combined into one which may cover the whole range of the Reynolds number. This composite formula is favourably compared with experimental data obtained previously by the present author.
    Download PDF (516K)
  • Kihachiro TAKETOMI
    1982 Volume 18 Issue 4 Pages 394-399
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    A new method of complex dielectric constant by using TE0ln mode in the cylindrical cavity resonator is suggested.
    The specimen is of the rod type and inserted by the micrometer drive along the center of a cavity resonator.
    A precise measurement formula of dielectric constant is given and accurate results are expected with the values in a formula are get from multi-points measurement.
    The method may be applicable to the specimen such as powder or liquid materials.
    Download PDF (727K)
  • Ken MISHIMA, Takao KUROKAWA, Hiroshi TAMURA
    1982 Volume 18 Issue 4 Pages 400-406
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    In this paper we analyze relations among the waveforms and the electromyograms (EMGs) of fast, or ballistic, movements of the forearm and propose a concept of the bimodal control policy of the motor center. The central commands described as change of impulse density in motor nerves are extracted from the EMGs on the basis that the ballistic portion of fast movements is solely under open loop control. Of the several parameters analyzed the amplitude of the agonist and antagonist commands and the time difference between the both command onsets closely correlate with the acceleration waveforms and they are suggested to be the essential central parameters that control the ballistic movements. In relatively small movements the command amplitude varies according to the movement amplitude while the onset time difference is kept rather constant. On the other hand, in larger movements the former takes the extremum value and the latter increases in proportion to the movement amplitude. These two exclusive ways of movement control are rational from the standpoint of attainment of high speed and accuracy of voluntary movements.
    Download PDF (1032K)
  • Keiichiro MIYATA, Hideo HANAFUSA
    1982 Volume 18 Issue 4 Pages 407-413
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    In the fluitlic digital circuits using reattachment amplifiers, pressure pulse signal is used usually. Pulse wave form varies with connecting conditions such as line length and branching. Prediction of the switching times of the amplifier for these inputs is very important in the circuit design.
    This paper first describes the switching model to calculate the distribution of the switching times of the reattachment amplifier, which is based on the transient response of an attached jet and the random switching characteristics. The jet displacement is obtained from the attached jet model which considers the control duct and the attachment bubble. It is assumed that the switching occurres due to the random switching and the switching probability is determined by the jet displacement. The effect of jet movement is included by an experimental constant.
    Next, the distributions of the switching times for the pulse inputs are measured to determine the experimental constant and the relation between input wave form and the distribution is discussed.Finally, the pressure wave forms and the distributions of the switching times are measured for the branched line system which is often used in the digital circuit. The characteristics of the wave forms are discussed and the effects on the distribution are made clear. The distributions for these inputs are calculated by using the switching model and the calculated results ware agreed with the experimental results. So it was verified that the distribution of the switching times could be calculated for the given input wave form by using this switching model.
    These results are useful for the design of the fluidic digital circuits using reattachment amplifiers.
    Download PDF (939K)
  • Shun-itsu HAMAO, Hiroshi HORIKAWA
    1982 Volume 18 Issue 4 Pages 414-419
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    This paper consists of three sections, the first section shows reciprocity problems for equivalent actuator circuits, the second section shows how to set up electric circuits having velocity feedback function without using velocity detectors. Lastly, a velocity feedback method is described, which has a new effect.
    There are two kinds of equivalent circuits for an actuator which converts electric energy into mechanical energy, such as voltage and current into force and velocity. These equivalent circuits differ in their reciprocity.
    Furthermore, the difference between reciprocity values for equivalent circuits depends upon physical actuator properties, such as those of electro-magnetic actuators and piezo-electric autuators.
    The first section reports the reason for the difference between the electro-magnetic actuator, investigated from the view point of the circuit network theory, and the piezo-electric actuator is shown as the inverse relation between the internal resistance of electric power source and the damping condition of a control system.
    The second section presents a construction method for an amplifier having the velocity feedback function without velocity detectors. This amplifier has the voltage feedback and current feedback function.
    However, as shown in electro-magnetic actuators and piezo-electric actuators, the relation between the positive feedback and the negative feedback in these control systems is reversed.
    The last section reports the two-terminalequivalent circuit expressing the control characteristics of a mechanical system, whose equivalent circuit always forms the reciprocity circuit. According to the study on this circuit, the elements that have been influenced by the velocity feedback are only electric circuit elements, not mechanical elements.
    This paper proposes a new velocity feedback method which is equivalent to adding a viscous damping factor to the actuator. The method will be useful for improvement in linear control system characteristics and for linearizing the system with static friction.
    Download PDF (896K)
  • Goro OBINATA, Hikaru INOOKA
    1982 Volume 18 Issue 4 Pages 420-422
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    Download PDF (423K)
  • Kensuke HASEGAWA, Takashi MIZUTANI
    1982 Volume 18 Issue 4 Pages 423-425
    Published: April 30, 1982
    Released on J-STAGE: March 27, 2009
    JOURNAL FREE ACCESS
    Download PDF (349K)
feedback
Top