All issues
Volume 20, Issue 4
Displaying 1-16 of 16 articles from this issue
- |<
- <
- 1
- >
- >|
-
Joe IMAE, Kyojiro HAKOMORI1984 Volume 20 Issue 4 Pages 281-287
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSMaximum Principle type necessary conditions are obtained for optimal control problems governed by discontinuous right-hand side state equations, without differentiability assumptions.
First, we define a new set-valued “Wrapper”, an extention of the concept of derivative, which is a good working tool for this kind of problems. Secondly, we construct a convex set similar to the cone of attainability proposed by Pontryagin et al. in the proof of their Maximum Principle. Lastly, we investigate certain properties of the set which play an important role in deriving the necessary conditions for the present problems.
Difficulties which arise from nondifferentiability assumptions and the existence of “separating hyperplanes” dividing the state space into a finite number of subregions are discussed in detail. The results presented here are considered to be an extension of those obtained by Masubuchi and Kanoh (1968) in the case where ordinary derivatives exist.View full abstractDownload PDF (906K) -
Toru FUJINAKA, Mitsuhiko ARAKI1984 Volume 20 Issue 4 Pages 288-293
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe optimal regulator problem of discrete-time systems which contain unilateral time delays is considered. Such systems appear when industrial plants with unilateral flow of materials are sampled. If the time delays are large compared with the sampling period, the computation to obtain the optimal law becomes very difficult due to the large dimensionality of the Riccati equation.
In this paper a method to reduce such computatuional difficulties is presented. First, the optimal regulator problem is solved for a smaller size imaginary system which is obtained by eliminating the time delays of the actual system. Then, the solution is substituted into the formula which yield the solution of the actual system with time delays. The method is applicable to systems with multiple time delays along the signal flow.View full abstractDownload PDF (602K) -
Takao HINAMOTO, Jiro SHIMONISHI, Sadao MAEKAWA1984 Volume 20 Issue 4 Pages 294-299
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe problem of observing the state of a two-dimensional system in the presence of unknown disturbance noise is considered. The Roesser state-space model is used to represent the two-dimensional system. Conditions are given for the existence of an observer capable of producing an asymptotic estimate of the state of the given system. Furthermore, conditions are presented for being able to choose the form of the observer in a simple form such that the transfer function has a separable denominator. Finally, design equations are developed for calculation of the observer matrices in the simple form where singular value decomposition is used to formulate the technique.View full abstractDownload PDF (688K) -
Yasuchika MORI, Takashi SHIGEMASA, Toshiyuki KITAMORI1984 Volume 20 Issue 4 Pages 300-306
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper describes a new design method applicable to multi-input/multi-output sampled-data decoupled PID and I-PD control systems, and its simulation test results. A special feature of this design method is that the sampling period can be selected independently for every control loop.
The systematic design method makes it possible to decide the control parameters for every control loop and to eliminate the interactions between control loops simultaneously after selecting a sampling period for every control loop. This method is a model matching scheme especially based on process low frequency characteristics, which can be identified with relative accuracy.
Rise-times for an individual subprocess involved in a multi-input/multi-output process are generally different from one another. In most conventional multi-input/multi-output design methods, it is common to chose the same sampling period in all control loops. In such a case, the control computer is forced to work rapidly, since all control loops must be controlled with the sampling period suitable to the shortest rise-time subprocess.
In the new design method proposed here, it becomes possible to reduce the large amount of data processing while satisfying the design requirements, because the sampling period can be selected considering its rise-time for every subprocess.View full abstractDownload PDF (824K) -
Katsunobu KONISHI, Toshio YOSHIMURA1984 Volume 20 Issue 4 Pages 307-313
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper is concerned with the minimum variance control of single-input single-output discrete-time linear stochastic systems with unknown constant parameters. A method is presented to stabilize the Self-Tuning Regulator (STR) given by Åström and Wittenmark, because the STR often causes the divergence or large peak value of the system output when the ratio of leading coefficient and its a priori estimate is not close to one.
Firstly, an admissible interval of the ratio for which the STR is stabilized is specified. Secondly, an identification algorithm of least squares type with two adjustable parameters (a. p.) is introduced, and the allowable region of the a. p. values is determined in order to ensure the global convergence of the corresponding adaptive control systems on the specified admissible interval. Lastly, a suboptimal a. p. value to suppress the output peak value is determined by maximizing a coefficient appearing in the equation used for convergence analysis. The effectiveness of the proposed method is discussed by the numerical examples in the fourth-order system.View full abstractDownload PDF (879K) -
Kinji MORI, Shyoji MIYAMOTO, Hirokazu IHARA1984 Volume 20 Issue 4 Pages 314-321
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSRecently distributed large scale control systems have been increasingly pursued and developed to attain not only fault tolerance but also on-line expandability and maintenability. In the distributed control system, a controller in a subsystem is required to have autonomy to continue its operation even after some faults occur in the system and control its own responsible object. Each controller reconfigures the system to stop the propagation of faults of the other controllers. Then the function of the system may be varied in accordance with the topology of the system which describes the connection among the controllers. Hence Functional Reliability is defined to relate function and reliability.
A function model of each controller is described as a event driven type function of outputs of the other controllers. A function attainability is defined as the possibility of coordination of functions among controllers. A functional reliability is defined as the mean functional attainability. This analysis method is the generalization of the conventional reliability analysis method.
This method is applied to design the system structure and the function allocation of the distributed control system in which each controller is connected by a loop network. This result shows the appropriateness of this analysis method to evaluate the distributed control system.View full abstractDownload PDF (1048K) -
Toshiro ONO, Haruo SHIMAOKA1984 Volume 20 Issue 4 Pages 322-328
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSQuickness and precision are the important properties which are required for an industrial scale used in production lines.
Remarkable progress of a microcomputer makes new approaches possible to meet the two contradictory requirements, and the approaches are realized through a set of softwares.
In this paper, we show“Dynamic Mass-measurement System of Displacement and Velocity Sensing Type”(DMS-DV), one of the best results of the new approaches, which consists of a spring scale, a microcomputer, and the displacement and velocity sensors. Then the software executed on the microcomputer plays an important role in the system.
The operation of the system is the following. Firstly, goods are put on the goods-plate of the spring scale, and the goods-plate is oscillated. Secondly, data of the displacement and velocity of the oscillation are processed in real-time by means of the software, and the mass of the goods is thus calculated.
In the data processing, the Practical DV-Algorithm is used, which is based on the idea of the dynamic measurement method. The special feature of the idea lies in the positive utilization of the dynamics of a measuring system.
The test system of the DMS-DV has been built up to examine the total performance and the real characteristic of the DMS-DV, though the feasibility of the practical DV-Algorithm was already verified by the computer simulation study.
The test system is a crude construction in comparison to industrial systems. However, the data obtained on the system are enough to verify the feasibility of the DMS-DV in a practical sense.View full abstractDownload PDF (1403K) -
Keisuke GOTO, Hideo TSUKUNE1984 Volume 20 Issue 4 Pages 329-336
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper describes a method for extracting elliptical figures in an image by the Hough transformation. We overcome the dimension problem inherent to the Hough transformation by decomposing the five-dimensional Hough transformation space into three sub-spaces based on the edge vector field properties of ellipses. A peak in a sub-space represents a group of ellipses which have the same geometric property. By sequentially selecting the peaks in the three sub-spaces we can extract elliptical figures.
Experiments using images of machine parts show that this method extracts ellipses of various shapes under practical conditions.View full abstractDownload PDF (2517K) -
Yukinori SUZUKI, Yoshinori MITAMURA, Hiroyuki YAMAMOTO, Tomohisa MIKAM ...1984 Volume 20 Issue 4 Pages 337-343
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe optimal driving condition of an assist pump (left ventricle-aorta bypass) applied to the cardiovascular system with a baroreceptor and periperal chemoreceptor loops was studied by computer simulation and animal experiments.
In the cardiovascular model, the peripheral system was represented by the Windkessel model and the heart by an elastance. Baroreceptor control loop regulates heart rate, cardiac contractility, venous compliance and arterial resistance. Peripheral chemoreceptor control loop regulates arterial resistance based on cardiac output. The assist pump was modeled by a variable compliance. Optimal driving condition minimizing tension-time index (TTI), maximizing diastolic pressure-time index/TTI ratio (EVR) or maximizing arterial pressure was analized.
Findings were: (1) Unloading of the left ventricle by an assist pump was augmented by cardiovascular control loops. (2) Optimal driving condition was different depending on with or without cardiovascular control loops. (3) Aortic pressure was not a suitable variable for estimating effectiveness of an assist pump, because it was kept almost constant by the cardiovascular control loops independently of driving conditions. (4) In the left ventricle-aorta bypass, TTI and EVR changed remarkably according to the driving conditions.
It can be concluded that cardiovadcular control loop must be considered when the optimal control system is designed for an assist pump.View full abstractDownload PDF (1065K) -
Mitsuji SAMPEI, Katsuhisa FURUTA1984 Volume 20 Issue 4 Pages 344-349
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper is to propose a method to control a mechanical system to move along laser beam, which may be useful for the teaching of the movement to robots. The mathematical model of the controlled system is constructed so that one of the coordinate corresponds to the laser beam, and the control problem is divided into the control along the beam and that of the deviation from the beam. Two controllers to track ramp are designed and practically realized by a minicomputer for these controls. They work satisfactorily when they are used for the control of three-dimensional linear motion mechanical system.
The sensor to measure the direction of and the deviation from the laser beam is also experimentally constructed and used for the experiment.View full abstractDownload PDF (687K) -
Yasuhiro OHYAMA, Osamu YAMANO, Katsuhisa FURUTA1984 Volume 20 Issue 4 Pages 350-356
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA Radio Control (RC) helicopter, which has various advantages due to small size and wireless operation, has not been used practically because of its difficulty in controlling its flight. In order to fly an RC helicopter desirably, some automatic flight control systems should be implemented.
In this paper, the mathematical model of a hovering helicopter is derived from the real helicopter model by approximating the rotor as one rigid stick and assuming that the direction of total rotor thrust is normal to tip-path-plane. The dynamics of two kinds of sensors, rate gyro and inclinometer are also investigated. Then the digital control system for the experimental RC helicopter mounted on gimbals is designed by using the above model, and it works satisfactorily. From the experimental results, the validity of the model and the control system is confirmed.View full abstractDownload PDF (790K) -
Yukio KIMURA, Sumio KOGA1984 Volume 20 Issue 4 Pages 357-360
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIt is well known that the faster runs a train, the more dissipates its energy. In train schedule the running time is settled in advance, so the train needs to be operated in fixed time between stations. As the train could adopt various operation patterns in the fixed schedule, the requisite energy for the train operation varies according to its patterns even in the fixed time and distance. We can, therefore, consider a train operation method which minimizes the requisite energy.
Our simple model of train operation enables to show that the train running energy is minimized when it runs in constant speed as long as possible in a specific time and distance supposing that the train moving energy were collected by the regenerating process in the brake operation.
While in case of no regeneration of the energy we gained another operation pattern of the minimum energy. The minimum energy has followed from the train operation that after keeping the constant train speed and coasting, the brake should be applied at the instant of a calculated speed when the train stops. But a commuters' train can not take the sufficient coasting time in it's frequently stopped and comparatively high speed operation. We have made clear the energy minimum operation of such a commuters' train as a special case of the no energy regeneration.View full abstractDownload PDF (553K) -
Case of Subsystem-State ObservationTsuneo YOSHIKAWA, Takashi IWASAKI1984 Volume 20 Issue 4 Pages 361-363
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSDownload PDF (304K) -
Tatsuya MASUDA, Katsuhiko FUJII1984 Volume 20 Issue 4 Pages 364-366
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSDownload PDF (354K) -
Hiroshi YAMADA1984 Volume 20 Issue 4 Pages 367-369
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSDownload PDF (253K) -
Tatsuo NARIKIYO, Masami ITO, Shigeyuki HOSOE1984 Volume 20 Issue 4 Pages 370-372
Published: April 30, 1984
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSDownload PDF (331K)
- |<
- <
- 1
- >
- >|