All issues
Volume 15, Issue 5
Displaying 1-25 of 25 articles from this issue
- |<
- <
- 1
- >
- >|
-
Katsuhisa FURUTA, Shigeyasu KAWAJI1979 Volume 15 Issue 5 Pages 565-571
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe compatibility of an observer to a plant is comprehensively discussed assuming that the plant changes its characteristics. The class of plants compatible to the given observer is characterized both in the case that the observer is used in the open-loop systems and in the case that it is used in the feedback systems as a regulator.
It is shown that the observer installed in the feedback systems is more robust than the one in the open-loop systems.View full abstractDownload PDF (884K) -
Kimio KANAI, Tadayasu DEGAWA1979 Volume 15 Issue 5 Pages 572-577
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSTwo different approaches to improve the convergence of transient characteristics are developed for the design of an adaptive observer for a single-input single-output linear discrete system where only the input and the output can be measured.
The first approach is to minimize the square function by a new recursive algorithm, in which the past data are exponentially weighted in order to emphasize the effect of current data. The second approach is to introduce unknown initial states into the recursive algorithm.
Simulation studies of a third-order system for different parameter selections are included at the end of the paper to demonstrate the effectiveness of the two schemes.View full abstractDownload PDF (676K) -
Kazufumi ITO, Sueo SUGIMOTO, Hiroshi SUGIYAMA1979 Volume 15 Issue 5 Pages 578-584
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe existence and uniqueness of the periodic nonnegative solution of the discrete matrix Riccati difference equations with periodic coefficients are discussed.
As is well known, such equations play a central role in the controls as well as filtering theory. The principal result obtained in this paper is that stabilizability and detectability is a necessary and sufficient condition for the existence of a unique nonnegative periodic solution to yield an equilibrium solution and for the associated closed-loop system to be asymptotically stable. This result corresponds to Hewer's result which was discussed with continuous periodic systems.
The proof of sufficiency follows along the line of quasilinearization technique and the necessity is shown by the effective use of the canonical structure theory which is developed to the discrete periodic systems in this paper. Also, for the discrete periodic systems the extended discussions on the definition of stabi-lizability and detectability by Hautas are included.View full abstractDownload PDF (927K) -
Masato TAJIMA, Kageo AKIZUKI1979 Volume 15 Issue 5 Pages 585-591
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper deals with a certain type of matrix Riccati equation with periodic coefficients, that is, a slightly generalized equation with the addition of a linear positive operator Π (t, ·) to the linear terms of the standard Riccati equation.
Such an equation arises, for example, in the quasi-steady state optimal control problem of a linear periodic system with a state-dependent noise, which is a natural generalization of the steady state optimal control problem of a linear time-invariant system.
In this paper, we discuss the existence of a periodic solution of the Riccati equation of this type and the asymptotic behavior of the solutions in connection with the periodic solution.
As a result, it is verified that under certain conditions of the coefficients (i. e.“stabilizability” and“detectability”), there exists a periodic solution to the above equation, if the coefficients Di(t), which form the positive operator Π (t, ·), are not too large. Moreover, we can see that under almost the same conditions, any solution of the Riccati equation converges to the above periodic solution in a certain sense as t goes to minus infinity. Then it is made sure that those results which have been obtained by G. A. Hewer et al. can be generalized to this case.
In the proofs quasi-linearlization methods and succesive approximation procedures are used effectively, so that the discussion becomes rather similar to the one which was adopted by W. M. Wonham.
Finally, we present one concrete example and apply the derived results to it.View full abstractDownload PDF (875K) -
Kazuo YAMANAKA, Etsujiro SHIMEMURA1979 Volume 15 Issue 5 Pages 592-596
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA stochastic stability problem for linear discrete-time systems with randomly varying delay is discussed. It is assumed that the delay is a discrete-valued stochastic process which is stationary and ergodic. The results concerning the stability are derived in connection with the stability of a system with a constant delay. Namely, it is shown that the system is almost surely asymptotically stable 1) if a specific value of the delay takes place with a sufficiently large probability and the system in which the delay is fixed to that value is asymptotically stable, or 2) if the delay alters itself with a sufficiently small probability and every system in which the delay is fixed to an arbitrary possible value is asymptotically stable.View full abstractDownload PDF (611K) -
Yoriaki BABA1979 Volume 15 Issue 5 Pages 597-602
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper deals with the relation between a sampling period T and the set of all initial states which are to be forced to the origin in less than m iterations with the period T by the admissible control. The set is called a controllable set and denoted by C(T, m). First, the author obtains some properties of the set C(T, m) for the 2nd-order single-input systems. And these results are extended to the nth-order r-input systems. Especially it is shown that the period p such that C(T, m)⊂C(p, m) is given by the equation p≥mT/(m-1) under some conditions irrespective of the system class.View full abstractDownload PDF (849K) -
Takahiko MORI, Masami ITO1979 Volume 15 Issue 5 Pages 603-608
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe generalized definitions of steady-state error for single input single output feedback systems in [6, Truxal] are extended to the case of multivariable (m input-m output) systems. And using this concept, a more practical definition of the system type of the steady-state error is presented. Also, it is shown that the system type studied so far in [2]-[5] is a special case of that given in this paper.View full abstractDownload PDF (641K) -
Hideji FUJIKAWA1979 Volume 15 Issue 5 Pages 609-615
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThough the state of the system with time-delay is infinite demensional, only a finite number of data can be handled in a practical calculation of the feedback control variable. Thus a problem of approximating an infinite dimensional state vector by a finite dimensional one, which preserves, in some sense, the essential information on the state, becomes important. In other words, it is a problem of the best approximation of a system with time-delay by a lumped parameter system. In this paper, the approximation of this sort by applying the method of weighted residuals is discusses. In order to measure the goodness of approximation, the location of eigenvalues and the shape of transient responses of the approximated lumped parameter system are compared to those of the original system with time-delay. From the numerical analysis on the several typical examples, it is shown that the method of weighted residuals gives results superior to those given by a commonly used simple difference method of the same approximation degree.View full abstractDownload PDF (833K) -
Shigenori MIZUGUCHI, Takeshi TSUCHIYA, Ryozaburo TAGAWA1979 Volume 15 Issue 5 Pages 616-621
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSAs method of parameter estimation for linear discrete-time systems in the presence of noise, the linear least-squares method based on the equation error is commonly used since the processing involved in the implementation is straightforward. As is known the method has a very serious drawback that the estimate is principally biased. In order to overcome this problem, recently, the generalized least-squares method and the instrumental variable method have been suggested. However, in these methods, it is not yet proved analytically that the estimates converge to the true parameter values. Also, in the generalized least-squares method, the noise parameters, which are not necessary if only the system is to be identified, have to be estimated in conjunction with the required system parameters.
This paper proposes a new linear parameter estimation method which is established with introducing the “Complex Equation Error.” The complex equation error corresponds to the Fourier transformation of the ordinary equation error. By applying this method, the consistent estimate of system parameters can be obtained through a simple linear procedure.View full abstractDownload PDF (696K) -
Hiroyuki TAMURA, Tadashi KONDO1979 Volume 15 Issue 5 Pages 622-627
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper deals with nonlinear statistical modeling of air pollution concentration for short-term prediction. The method used for modeling is a revised version of the basic GMDH (Group Method of Data Handling) algorithm. By using the time series data of the SO2 concentration, the wind velocity and the wind direction in Tokushima, Japan, we intend to find a suitable model for predicting SO2 concentration a few hours in advance. Firstly, the suitable data length for modeling air pollution in Tokushima is investigated. Secondly, three different prediction models obtained by the revised GMDH are compared to find a suitable structure and suitable input variables in the model. The predicted results obtained by the revised GMDH model are compared with the results obtained by a linear regression model, a linear autoregressive model and a basic GMDH model. It is shown that the revised GMDH model developed in this paper gives a better result and compares favorably with linear models and the basic GMDH model. It is also shown that the revised GMDH model obtained is much simpler than the basic GMDH model.View full abstractDownload PDF (735K) -
Koji TAJIMA1979 Volume 15 Issue 5 Pages 628-633
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSTo identify a multivariable stochastic linear system with unknown structure, we suggest the use of a particular vector difference equation model. This model requires only one structure index (minimal order) for its unique representation. Thus it is not necessary for consistent parameter estimation of this model to determine the canonical structure indices. A new method proposed for determining the minimal order of this model from input-output observations is based on the behavior of the estimated Markov parameters. By using this method, we can determine the model order with less computational burdens than the conventional methods. Simulation results are presented to illustrate the model order determination method.View full abstractDownload PDF (668K) -
Tetsuo MABUCHI, Toyomi OHTA1979 Volume 15 Issue 5 Pages 634-640
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA decomposition method may be classified as near-optimization method of linear regulators with small parameters.
To increase the accuracy of approximation, it might be recommended to increase the order of the unperturbed system; but it will require a tedious and laborious calculation.
A method, the so-called “repeated-decomposition” is developed to circumvent the difficulty.
This is composed of iterations of decompositions in the expanded unperturbed systems. The repeated-decomposition method will have less calculation keeping the same degree of approximation.View full abstractDownload PDF (865K) -
Covariance Specification for Linear Time-Invariant Feedback Control System with Colored System NoiseMasahiko NAGAYASU1979 Volume 15 Issue 5 Pages 641-646
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper describes the covariance specification problem for the linear time-invariant state-feedback control system disturbed by a colored system noise. By definition the covariance specification problem is to find the state-feedback control with which the stationary state covariance matrix coincides with the prescribed covariance matrix. The author solved a covariance specification problem for a system disturbed by a white system noise before.
In this paper the problem is extended to a system disturbed by a colored system noise. The necessary conditions of the optimal feedback gains which make the resulting state covariance coincide with the prescribed one under the minimum control effort are derived, and a sequential substitution algorithm is given for a numerical solution.
As a numerical example the control problem of an aircraft in the atmospheric turbulence is discussed.View full abstractDownload PDF (789K) -
Takeshi YAMADA, Hidenori KIMURA1979 Volume 15 Issue 5 Pages 647-653
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe main function of the control system of a tandem cold-rolling mill is to produce the strip with a prescribed thickness accuracy and to maintain the interstand tensions within a a reasonable range in the rolling operation, against disturbances such as the mill input thickness, the stand set-up error, the drift of ingoing strip hardness, the coefficient of friction and so on. In a conventional control scheme, the output thickness is controlled by adjusting the screw position, while the interstand tension is controlled by updating the roll speed. Since the output thickness is strongly coupled with interstand tension, it is natural to consider that the integration of thickness control and tension control will result in a better control performance.
In this paper we demonstrate an integrated control scheme for output thickness and interstand tension using the modern state-space approach. A state-space model for 5-stand cold-rolling mill is derived and an optimal feedback control law is calculated for this model, incorporated with an integral action to eliminate the tension off-set. The output thickness of the preceding stand is employed as a feedforward signal to update the screw position of each stand.
Simulation results are exhibited in comparison with that of a conventional control scheme. It is concluded that our system demonstrates a better control performance over the conventional one. Moreover it suggests an approach for shape control by adjusting weighting matrices of the performance index.View full abstractDownload PDF (705K) -
Hironori HIRATA1979 Volume 15 Issue 5 Pages 654-659
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe distribution of roots of a random algebraic equation, whose coefficients are dependent on one another, is studied. Each coefficient is assumed to take values of a normal distribution with an arbitrary mean and an arbitrary variance. In addition, the relation between the distribution of roots and the degree n of the algebraic equation, i.e., how the complex roots distribute as n→∞, is discussed.View full abstractDownload PDF (570K) -
Hiroshi NAGANO, Minoru HIGASHIGUCHI1979 Volume 15 Issue 5 Pages 660-665
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe common design techniques of control systems are based on mathematical descriptions of the system to be controlled (i. e. the plant). Particularly, linear models are used in many cases because the plant can be almost always assumed to be linear for small signal and because linear models give us a good prospect for designing control systems, even though we must consider nonlinear characteristics of the plant at the last stage of designing.
Making a linear model necessarily means to replace uncertain parameters with constants. Such a “constant” may represent a random variable, a membership of a fuzzy set, a time-varying function or a state-dependent variable.
Then we need to evaluate the influence of the “constant” parameter's perturbation on the design specification of the control system.
For this purpose, sensitivity analysis and statistical testing have been used. But they deal with a constant perturbation of the “constant” parameter and they are not powerful enough for all cases stated above.
In this paper, we evaluate the upper and lower bounds of the influence of the “constant” parameter's perturbation on the design specification of the control system when the parameter value is bounded. Consequently, it becomes clear that the influence of the perturbation in this sense is several times as much as that of a constant perturbation.View full abstractDownload PDF (705K) -
Haruhiko ASADA, Hideo HANAFUSA1979 Volume 15 Issue 5 Pages 666-671
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper describes the handling of objects by artificial fingers, when they include some extra fingers which are not always needed in a simple gripping. Firstly, the number of controllable finger forces and the handling forces to manipulate objects are compared, and the redundancy of the finger control system is defined. Next when the redundant fingers are contained, there exists some freedom to select the finger forces which generate the same handling force. We derive the optimal solution of finger forces by using pseudo inverse matrices. Finally, the experiments verify the theoretical results and the utility of the redundant fingers to handle objects.View full abstractDownload PDF (1456K) -
Hiroshi SAITO, Hideki UCHIYAMA1979 Volume 15 Issue 5 Pages 672-677
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe amount of dissolved oxygen (DO) in water which is one of the important indices of water pollution is ordinarily determined by means of Winkler's method. Because of the manual analysis, however, the errors are liable to occur in the process of pre-treatment of sample or in the visual detection of a titration end point.
A new analyzing instrument which controls automatically the sample treatment and titration procedure was constructed to determine the amount of DO more accurately. The instrument consists of the following parts: devices to measure the volume of sample, to treat the sample, to titrate, to indicate the result and to control the system. By the use of a potentio-metric method with a constant current for detecting a titration end point, the construction of the detector is considerably simplified. A series of the measurement is performed according to a pre-determined program. A digital indicator gives the results of the analysis, that is, the concentration of DO and the volume of titrant used for titration.
The precision of the analysis is experimentally evaluated to be about ±0.02g/m3 in terms of the standard deviation over the concentration range of DO from 8.5 to 13.5g/m3 and no significant difference is observed between the results obtained by this method and by the manual Winkler's method.View full abstractDownload PDF (954K) -
Teruo ARAI1979 Volume 15 Issue 5 Pages 678-682
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA number of gypsum specimens, prepared by deposition of plaster of Paris (CaSO4 0.5H2O) in water at 1:3 mass ratio, are studied in view of their suitability for use as diffuse reflectance standards in such field work as the remote-sensing from the aircraft.
Comparison of their spectral reflectances and goniophotometric characteristics with those of some samples of other materials revealed the following advantages of the gypsum; 1. high reflectance, 2. fairly flat spectrum, 3. nearly perfect diffuseness, 4. good reproducibility of characteristics and 5. easiness of access, preparation and handling.
In connection, gypsum mixed with a little amount of India ink is studied in the same manner and found to be useful and convenient as a gray reflectance standard.View full abstractDownload PDF (1272K) -
Masakazu HARADA, Osamu OYAMA, Shinya YUASA1979 Volume 15 Issue 5 Pages 683-688
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSWhen the pressure of control signal increases gradually across the switching level of a device, the device switches with varying switching delay. The combination of restrictor and solid capacitor gives ±5% to ±7% in accuracy. To improve the accuracy of a pneumatic delay, it is necessary that the control signal rises rapidly when crossing the switching level of the pneumatic or fluidic device.
To realize the rapid rise, a new delay element using a thin film was developed. A film bag is installed in a chamber with bleed holes. Initially the volume must be reduced to zero. When the signal is turned on, the flow starts to expand the film, but the pressure does not rise until the film is inflated fully in the chamber.
The histograms of the new delay element have shapes similar to those of Gaussian distribution, and the ratio of ±2σn to the mean delay may be regarded as an index of accuracy. The effects of the restrictor, the pressure supply and the type of devices are investigated, and the delay element with the film type capacitor can realize ±0.4% to ±0.7% accuracy.View full abstractDownload PDF (776K) -
Makoto IBUKA, Haruo HOSOMATSU1979 Volume 15 Issue 5 Pages 689-694
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper describes a new type of thin film Superconducting Quantum Interference Devices, (SQUID). SQUID is a very sensitive magnetic sensor whose principle of operation is based on the magnetic flux quantization and the Josephson effects.
After a brief summary of SQUID operation, the construction and properties of Nb thin film SQUID are dealt with in some detail. This includes the production process and sensitivity estimates.
An Nb thin film is deposited by rf-sputtering on a flat substrate and a microbridge is fabricated with photolithography, sputter-etching and oxidization techniques. This structure is suitable for massproduction and has an excellent mechanical stability. Several SQUID stored at room temperature for three years during which they were periodically subjected too liquid-helium temperature more than 50 times showed no significant deterioration.
The last part of this report covers the original technique used to lower SQUID ring inductance and hence increase sensivity to obtain a resolution better than 4×10-19Wb/√Hz.View full abstractDownload PDF (2431K) -
Toshiyuki KITAMORI1979 Volume 15 Issue 5 Pages 695-700
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA design method will be useful if it calls for a minimum amount of knowledge about the controlled process, since identification of the state equation is not an easy task especially in the field of process control. The author previously developed a method for the continuous-time case which involves only a limited number of the controlled process parameters corresponding to the number of the controller parameters. In this paper the method is applied to sampled-data control systems.
The controlled process is usually time-continuous even in a sampled-data control system. Thus, the system is working on the continuous-time basis. Here, sampled-data control systems are approximated into continuous-time ones, and the previously developed method is employed to give the design formulas.
The control system to be designed is to meet the following specifications:
(C1) It has zero steady-state error.
(C2) It has adequate damping characteristics.
(C3) It has the shortest rise-time on satisfying (C1) and (C2).
The design formula gives rise to a sampled-data system which approaches to a continuous-time one with an increasing response speed but without changing its damping characteristics as the sampling period reduces to zero. Zero sampling period reduces the formulas to the previously obtained ones for the continuous-time systems.
Formulas for the PID and the I-PD control schemes are derived according to the method. Some designed samples are demonstrated in step response curves, which show effectiveness of the method. The sampling period T is assumed in advance to be less than half the rise-time of the controlled process, otherwise a system with T=0, that is, a continuous-time system will result because of the specification (C3) above.View full abstractDownload PDF (691K) -
Toyohiko SHIMIZU, Masami ITO1979 Volume 15 Issue 5 Pages 701-702
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSDownload PDF (232K) -
Shigeyuki HOSOE, Yutaka KANEDA, Masami ITO1979 Volume 15 Issue 5 Pages 703-704
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSDownload PDF (256K) -
Akitoshi KAWAMURA, Kunihiro HAYASHI, Susumu HATTORI1979 Volume 15 Issue 5 Pages 705-706
Published: September 30, 1979
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSDownload PDF (236K)
- |<
- <
- 1
- >
- >|