All issues
Volume 20, Issue 7
Displaying 1-17 of 17 articles from this issue
- |<
- <
- 1
- >
- >|
-
Setsuo SAGARA, Takami MATSUO, Kazushi NAKANO1984 Volume 20 Issue 7 Pages 573-578
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn a deterministic system of which the inputs or outputs are detected with errors, the observer designed without considering the errors does not often give the efficient estimates of the states. Therefore it is necessary to construct the observer which guarantees the efficient estimates notwithstanding the errors.
Komiya and Furuta defined a bias free observer as the observer effective for the output biases which are governed by the known linear dynamical system with the unknown initial conditions, and designed it by geometric approach. But their method cannot be used in the case of such a more general class of detection errors than the biases as given by outputs of the unknown dynamical system.
In this paper we propose a design method of the detection error free observer (DEFO) by spectral decompostion approach when the detection errors belonging to C1-class act at the outputs. The necessary and sufficient conditions are given to construct the DEFO for estimating the complete states. Furthermore, the sufficient conditions are presented to construct the DEFO for estimating the partial states. Finally, we show numerical examples of the proposed DEFO.View full abstractDownload PDF (673K) -
Application of Finite Time Settling ControlKenzou WADA, Nobuo HAYANO, Tomosada JYOTAKI1984 Volume 20 Issue 7 Pages 579-586
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this report, we discuss the weight control of hopper by using a microcomputer.
The control processes in which the granular material is concerned are generally difficult, because the granular material reveals a very complicated nature in its handling.
In the present investigation, we adopted a digital controller with a more elabolate control action than that of an analog controller. At the first step, we used sand as granular material due to its good flowability. As for the control algorithm, we used the “Finite Time Settling Control (F.T.S.C.)” method that makes a very excellent response for the step change of a set point value and an external disturbance, respectively. We estimated the unmeasurable state variables by using “Finite Time Settling Observer (F.T.S.O.)” or “Minimal Order Observer (M.O.O.)”.
The following results became clear through the experiments and the simulations by using a microcomputer.
1) If we use F.T.S.C. method as the control algorithm, P-action is more suitable than I-action for the process containing a saturation element.
2) In our system, in the case of P-action, there is an off-set for the step change. To avoid this off-set, the set point value must be adjusted prior to the experiment. Contrarily, in the case of the external disturbance input, if we use F.T.S.O. as the observer in the system, the off-set disappears.
3) In the case using M.O.O., we could eliminate the off-set by the suitable choice of the parameter of the M.O.O. itself and the system gains.View full abstractDownload PDF (1693K) -
Kozo OKAZAKI, Shinichi TAMURA1984 Volume 20 Issue 7 Pages 587-591
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSWe describe here an approach for identifying the transfer function for the continuous-time model of a system. The correlation method in this paper is as follows; the cross-correlation functions (CCFs) between the M-sequence Gaussian (shape) pulse (MG) signal and the n-th differenciated value (nD) of the M-sequence input signal, and the CCFs between the MG signal and the nD of the system output are obtained by the CCFs between the nD of the MG (nD-MG) signal and the M-sequence input signal, and the CCFs between the nD-MG signal and the system output respectively. By this method; (1) we can get an un-biased estimation in a white observation noise system. (2) The method can be applied off-line or on-line (non-recursively or recursively). (3) In comparison to the two-stage least-squares method and the instrumental variables method about the correction of the bias caused by the noise, the described method filters off the noise in the first place using the MG signals; then the estimation is obtained by the least-square method. In addition, it is shown by the experimental results that even if the transfer function of the model is different from that of the system, it is possible to obtain the system parameters in a reasonable form.
Our experimental results of digital simulation demonstrate that the proposed method can identify the system fairly well.View full abstractDownload PDF (608K) -
Takashi SHIGEMASA, Yoshinori ICHIKAWA1984 Volume 20 Issue 7 Pages 592-599
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper proposes a new auto tuning method for digital process control system under closed loop operation. The auto-tuning method is composed by combining an online parameter estimation method under the closed loop identifiability conditions, a newly developed conversion method from the identified pulse transfer function into continuous time transfer function and a digital control system design method based on partial model matching principle.
Based on the auto-tuning method mentioned above, a practical auto-tuning algorithm for digital PID process control system with reference input filter is shown. Moreover, the practical transfer function conversion algorithm newly developed is described in detail.
Response curves, which are demonstrated in experimental study, show the effectiveness of the proposed auto-tuning method.View full abstractDownload PDF (1064K) -
Yukio NISHIMURA, Kazunobu YOSHIDA, Yoo YONEZAWA1984 Volume 20 Issue 7 Pages 600-605
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper the design problem of a linear system under input and/or state constraints is investigated. That is, in the control system x=(A-bk')x=A(k)x, find the optimal feedback gain which minimizes the system performance index, under state constraint such as|l'x|≤1 and initial states as x0∈Ω0 (bounded and closed set). In problems of this kind, in general, evaluations of many trajectries starting from various initial states are necessary so far to know whether the gain k is admissible in the sense that given inequality constraint is always met. A simple criterion is proposed here which permits us to know whether a gain k is admissible or not by calculating only one trajectory. That is, it is proved that the necessary and sufficient condition for a gain k being admissible is z(t)∈Ω0 for all t≥0, where z(t) is the solution of the reversed-time adjoint system z=A'(k)z, z(0)=l, and Ω0={z||z'x0|≤1, x0∈Ω0}. Next, the expression of the gradient of maximum value of constrained variable with respect to gain k is led, and using this a design algorithm is proposed by which optimal feedback gain is obtained. A sufficient condition that is to use to give efficient criterion for admissible gain is also presented. A design example of a tank system with some constraints is shown to demonstrate the usefullness of proposed method.View full abstractDownload PDF (674K) -
Yoichi HAYASHI, Naonori NISHIDA1984 Volume 20 Issue 7 Pages 606-613
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSMany important problems can be represented as dynamic network flow problems. The dynamic networks are used for the analysis of transportation systems, material handling systems, network models for building evacuation, etc.
Minieka has presented an algorithm based on Ford and Fulkerson maximal dynamic flow algorithm that constructs a maximal dynamic flow with a latest departure and an earliest arrival schedule.
Furthermore, the problem of finding a dynamic flow with desired distribution of departures and arrivals was considered by the present author previously. The dynamic network has a single source and a single sink, and further, holdover nodes.
In the present paper, we extend the problems stated above to the case for the dynamic networks with multiple sources and sinks, which include holdover nodes of different properties. This problem is applied to the network model for building evacuation. Since the evacuation of a building involves the flow of people through well defined passageways, it is natural to consider the evacuation problem to be dynamic network flow problem.
We consider the problem of finding an evacuation route which realizes the desired pattern of evacuees through the exits. The properties of the dynamic network are also considered by using the classification of holdover nodes. Illustraive examples for the evacuation of a three floor building are solved to demonstrate the effectiveness of the proposed methods.View full abstractDownload PDF (1121K) -
Kaname YOSHIDA1984 Volume 20 Issue 7 Pages 614-620
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper concerns with the decomposability in the approximate optimization method of geometric programming problem.
There are some cases of that we need only an approximate optimum of the given problem with a little caluculation. The approximate optimization is proposed to satisfy such the request. This method is especially effective for the problem which can be decomposed into some smaller scale problems whose degree of difficulties are all zero. But it is not easy to check whether the sufficient condition which has already been shown to decompose the given problem into the zero degree of difficulty problem is satisfied or not. So, the checking the condition has been a stumbling block, when we apply the approximate optimization method to the given geometric programming problem.
Here we show a class of the problem which can be decomposed in to some smaller scale problems whose degree of difficulties are all zero. The condition that the problem belongs to the class is discribed with simple form such that F=d-1(0)+1 where d is a number of degree of difficulty and 1(0) is a number of the terms inclue ded in the objective function. If the sign of F is not positive, then the problem can be decomposed in to the zero degree of difficulty problems with probability 1. As the condition can be easily checked, we can use the approximate optimization method to many problems. Last we apply this result to the ship operation problem and sea power system design.View full abstractDownload PDF (798K) -
Hideyuki SUGIYAMA, Kozo MATSUSHIMA1984 Volume 20 Issue 7 Pages 621-626
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, we propose a new method by means of fuzzy model for the analysis of the human operator's property in a manual control system and the availability of this method is verified referred to the controlled element which is a servomechanism with nonlinearity of torque saturation.
We investigate operator's control behavior for a step input and construct its fuzzy model in terms of the control deviation and the velocity of controlled variable with the assumption of a discrete process of operator's inference.
As a result, fuzzy model is found useful to simulate and to analyze the human operator's control behavior in man-machine systems.View full abstractDownload PDF (730K) -
Ario OHSATO, Takashi SEKIGUCHI1984 Volume 20 Issue 7 Pages 627-634
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper presents a method for solving the convexly combined form of composite fuzzy relation equations, λ·(x_??_R)+λ·(x_??_R)=y, for the unknown two fuzzy relations R, R and unknown fuzzy convex combinator λ knowing two fuzzy sets x and y. It is another aim of the paper to show that the convexly combined form of composite fuzzy relation equations makes it possible to formalize a corresponding relation between any arbitrary two fuzzy sets x and y. Furthermore, it is suggested that the convexly combined form of composite fuzzy relation equations and the method of solution will provide a fundamental strategy to solve the identification problem of fuzzy input-output systems.View full abstractDownload PDF (1041K) -
Ichiro KIMURA, Yoshio NAKAMURA1984 Volume 20 Issue 7 Pages 635-640
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSDevelopment of a new temperature-sensing actuator with ferrite and a rare-earth cobalt magnet for low cost and simple instrumentation was reported by the authors on the No. 9 Vol. 18 of this transaction.
The performance of this device is limited by the difference of the squares of the magnetic flux density at the air-gaps at high temperature and at low temperature, on the condition that in both cases the moving iron is hold apart from the magnetic poles.
The magnetic flux distribution has been investigated experimentally, and on the basis of those experimental results, the yokes and the moving iron have been redesigned as follows.
(1) The shape of the ferrite piece has been simplified from the letter “inverse T” (10mm long) to a simple cylinder of the same length. This simplification makes working easier.
(2) The shape of the moving iron has been changed from the letter “U” to a flat type.
As a result, the flux density at the air-gaps in the case of wide air-gaps at high temperature has been increased.
Thus, while the size and the weight of the device have been decreased the pull force vs. temperature characteristic has been still improved. The net actuating force per unit weight of the device has been increased from 0.83N to 1.3N each per 10g.
Further, several properties which should be considered in order to put this device to practical use, such as response time, influence of vibrations and influence of ferromagnetic materials, have been tested and its results are reported in this paper.View full abstractDownload PDF (791K) -
Naoyuki TAMARU, Akira NAGAYAMA1984 Volume 20 Issue 7 Pages 641-647
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn a magnetic tape unit, high speed and precise positioning control should be achieved. This paper demonstrates that a Kalman filter can precisely estimate the state variables for the capstan motor, combined with its torsion resonance characteristics modeling, and widen the loop bandwidth by the feedback from estimated state variables.
The capstan motor is modeled as a two freedom resonant system that consists of two parts, a torque generator part and a capstan part. The Kalman filter is constructed to estimate rotation angle and rotation angle velocity for the two parts, from the motor driving current and the capstan rotation angle.
The comparison between experimental frequency response for estimated state variables and calculated values confirmed that the two freedom model for the capstan motor was able to describe its actual resonant characteristics and that the Kalman filter circuit operated up to several kHz beyond the first resonant frequency for the system. The actual capstan motor was controlled by feedback from these estimated signals, based on the optimal control theory. The result shows that the loop bandwidth for the positioning control was extended about twice as wide as the conventional method, where a tachogenerator and band reject filters were used. The motor was started up for about 0.4ms.View full abstractDownload PDF (1465K) -
Ato KITAGAWA, Toshiharu KAGAWA, Toshio TAKENAKA1984 Volume 20 Issue 7 Pages 648-653
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSAn accurate and efficient approximating method is developed for simulating a transient laminar flow of pneumatic transmission line to be analyzed taking into account the frequency-dependent friction and the heat transfer by the method of characteristics.
This method consists of approximating the exact W. Zielke's weighting function about the wall shear stress and the exact F. T. Brown's weighting function about the heat transfer. Each of them is expressed by a sum of impulse responses of first order time lag systems.
The calculated step response of blocked line shows a good agreement with the experimental result.
By introducing dimensionless variables for the basic equations, dimensionless parameters are induced, which describe the transient response. The step response of blocked line is expressed by using these dimensionless parameters.
This method requires no large computer storage capacity, shortens drastically the computation time and produces no unacceptable error due to approximation.View full abstractDownload PDF (640K) -
Kazuhiko KUDO, Hozumi USUI, Masao OHTA, Yoshikuni SHINOHARA1984 Volume 20 Issue 7 Pages 654-659
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA hybrid simulator was made to calculate the dynamics of high temperature gas cooled reactor (VHTR). The continuous space-discrete time (CSDT) method is applied to solve the partial differential equations of the heat transfer in the hybrid computation. By this method the error of the heat balance is decreased to less than one percent in the steady state. Though the mini computer is used for this simulator, it operates about five times faster than real time.
The dynamics of VHTR are characterized by the large heat capacity of the reactor core and the long time constant. The values of these parameters are reported as the results of this calculation. The control system of the reactivity and the coolant flow rate is required to operate the reactor. The nonlinearity of VHTR which occurs in the change of flow rate are also understood quantitatively by this simulator.View full abstractDownload PDF (891K) -
Shogo TANAKA1984 Volume 20 Issue 7 Pages 660-662
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSDownload PDF (393K) -
Position ControlKatsutoshi TAMURA, Kunihiko ICHIKAWA, Kenji KOTAKI, Yasushi AMANO1984 Volume 20 Issue 7 Pages 663-665
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSDownload PDF (462K) -
Yoshinori ANDO, Masayuki SUZUKI1984 Volume 20 Issue 7 Pages 666-667
Published: July 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSDownload PDF (246K) -
1984 Volume 20 Issue 7 Pages 669
Published: 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSDownload PDF (139K)
- |<
- <
- 1
- >
- >|