巻号一覧
12 巻 (1976)
- 6 号 p. 613-
- 5 号 p. 491-
- 4 号 p. 367-
- 3 号 p. 239-
- 2 号 p. 121-
- 1 号 p. 1-
12 巻, 5 号
選択された号の論文の18件中1～18を表示しています
- |<
- <
- 1
- >
- >|
-
古田 勝久, 森 正三, 原 辰次1976 年 12 巻 5 号 p. 491-496
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーThe compatibility of Luenberger observer designed for a time-invariant linear system is comprehensively discussed and an explicit statement is given for a class of systems to which a given observer is compatible. This class indicates allowable variations of system parameters such that the designed observer functions as an observer even when the parameters are changed. The existence condition and the design method for the observer which has zero sensitivity property to plant parameters are described. Further, as an application, the zero sensitivity of a regulator combined with an observer is shown for a class of parameter variations.抄録全体を表示PDF形式でダウンロード (726K) -
川路 茂保, 手嶋 俊明, 遠山 栄二1976 年 12 巻 5 号 p. 497-503
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーAn optimal linear regulator calls for feedback of all the state variables in the plant, but they are not accessible for direct measurement under many realistic situations.
This paper is concerned with a design of the dynamic controller, which is constructed by a linear combination of the outputs of the plant and the observer with an order lower than those of a state observer or a functional observer.
The design parameters are chosen to minimize the expectation of the regulator cost functional when the mean and covariance of the initial state are known, or a ratio of an increased cost to optimal cost as the index of goodness of observer is adopted when no priori information of the initial state is available.
The design procedure and an example are shown for a single output system, and the results are compared with those of a state observer or a functional observer.抄録全体を表示PDF形式でダウンロード (891K) -
須田 信英, 藤井 隆雄, 山内 良一1976 年 12 巻 5 号 p. 504-510
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーThe problem of performance deterioration in optimal regulators is discussed for linear discrete-time systems with Kalman filters or minimal-order state observers as state estimators.
First, necessary and sufficient conditions are derived for which we can choose the design parameters of the observer in order to make the performance deterioration arbitrarily small or equal to its minimal value. Similar design problems of the observer are investigated also when these conditions are not satisfied.
Secondly by comparing the above conditions for these two types of observers, it is shown that the flexible minimal-order state observers are superior to Kalman filters.
Finally the results obtained here are compared with those obtained for continuous-time systems.抄録全体を表示PDF形式でダウンロード (1012K) -
石原 弘一, 添田 喬1976 年 12 巻 5 号 p. 511-517
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーThis paper deals with the problems of reconstruction and the observation method for continuous time linear dynamical systems with random parameters. The Markov model of stochastic differential equation is used to develop the physical continuous systems.
The discussion of the problem is mainly divided into three parts. Firstly, the unknown statistics of the state in random parameter systems subjected to deterministic and piecewise continuous inputs are reconstructed from the system's scalar output at n(n+1)/2 points in its history.
Secondly, the conditions of reconstructibility for the unknown stationary or non-stationary random inputs to the random parameter systems with known initial states are considered from the view-points of input observability.
Finally, an optimal observation matrix and optimal monitor time locations are determined from the min-max procedure on the eigen value of the symmetic matrix relative to the observability matrix, which is induced by the condition to minimize the difference between sample and population statistics in the sense of norm.
An numerical example is given to illustrate the obtained conditions and the proposed technique. The relations between the confidence intervals and the observation method are discussed in detail.抄録全体を表示PDF形式でダウンロード (942K) -
太田 裕文1976 年 12 巻 5 号 p. 518-523
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーCurran's model matching method is outlined and refined in this paper. The main points of improvement are: to assure the nonsingularity of the matrix L1 so that the algorithm may always work; to show the applicability of the algorithm to the systems in which the input terms are included in the output equations; and also to the systems with different input/output dimensions. The former two points are brought about by utilizing the equicontrollable canonical form representations with augmented controllability index. These results enable us to complete the matching, if the plant and model have the same number of inputs and outputs, within the allowance of the modification by the factor, βν/(s+β)ν, which is introduced to the model in order to assure the realizability of the dynamic compensator. It is also shown that the stability property of the compensator discussed by Curran is false. The efficiency of the algorithm will be illustrated in the numerical examples, and the effects of the location of the unobservable poles on the performance of the augmented plant will also be discussed.抄録全体を表示PDF形式でダウンロード (774K) -
小柳 孝巳, 愛田 一雄, 河田 祥志, 泉 久信1976 年 12 巻 5 号 p. 524-529
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーIn this paper, two kinds of methods are proposed to decrease the effects of disturbances in the identification of process dynamics of an adaptive control system using the m-sequence signal.
In order to eliminate the effects of low frequency disturbances, such as drift, added to the system output, a new digital correlator is presented based on the calculating principle of the cross-correlation function φxy(τ)=x(t)y(t+τ).
To minimize the effects of wide bandwidth disturbances, exp (πζ/√1-ζ2) is used as a performance index for adaptive control and is calculated from the ratio of the first plus area and the first minus area of the impulse response of the control system to be identified.
This method has also the following merits.
(1) The method is less affected by the bias error of cross-correlation function.
(2) The time required for identification is reduced.
(3) The number of integrators of the correlator is reduced.
Moreover, it is reported that the error of estimation of the index performance can be reduced by correcting the bias error of crosscorrelation function.抄録全体を表示PDF形式でダウンロード (860K) -
中溝 高好, 大城 正和1976 年 12 巻 5 号 p. 530-535
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーThe partially singular stochastic regulator problem is solved in this paper for a time-invariant linear system with noisy observation. The performance index is taken here as J=E{x'Qx+u'Ru}, in which the control weighting matrix R is not non-singular. The explicit form of the optimal dynamic controller is first derived. It is then shown that the dimension of the optimal dynamic controller is lower than that of a non-singular case. The singular optimal filtering problem in which the covariance matrix of the observation noise is singular is also briefly discussed, and the duality between the both problems is demonstrated.抄録全体を表示PDF形式でダウンロード (719K) -
花房 秀郎, 浅田 春比古1976 年 12 巻 5 号 p. 536-542
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーIt is desired that artificial fingers can grip various objects with a gripping suitable to their shapes. In this paper, we discuss the matching between the structure of artificial fingers and the shapes of objects. The gripping by a robot is regarded as the restriction of the objects by the fingers. The forces acting on the objects should be balanced for the gripping.
First the finger forces depending on the structure of the artificial fingers and the shape of an object are analyzed. Second by introducing the potential function the stability of a gripping form is discussed. When the potential is relative minimum at a certain gripping form, the resultant of all the finger forces acts so as to keep the gripping for all of small deviations. The defined potential can be applied to decide the optimal gripping form and the design of an universal gripper.抄録全体を表示PDF形式でダウンロード (833K) -
広瀬 茂男, 梅谷 陽二1976 年 12 巻 5 号 p. 543-547
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーA biomechanical study of living systems suggests us some interesting mechanisms which enable an organism to carry out specific functions. Among various organisms, the authors deal with the organism having a slender configuration and making active and flexible winding motions with muscles attached along the body. In spite of the simple shape, it performs miscellaneous functions which have never been realized in an artificial mechanism.
The authors call this type of organism as Active Cord-Mechanism (abbreviated as ACM) and have been studying its motion, control and other problems associated with an engineering application.
The basic locomotive motion of ACM on an even ground has already been realized in a snake-like shape.
In this paper, some of the problems to develope this locomotive ACM into a more intelligent machine are discussed.
First, the installation of tactile sensors is discussed and the biological principle of lateral inhibition is shown to be useful for controlling the ACM with tactile sensors.
Second, the efficient and smooth controlling principle is proposed which the authors call the modified first order data hold.
Finally, by the ACM mechanical model with tactile sensors and reconstructed controller, more intelligent locomotive experiments have been made, to demonstrate that the locomotion by detecting the obstacle in labyrinth, or the propulsion by pushing a fixed object have been successfully accomplished.
Thus the authors believe that this study solves some of the controlling problems of locomotive ACM to develope as an intelligent machine.抄録全体を表示PDF形式でダウンロード (1545K) -
Al蒸着面上のコレステリック液晶膜の場合田村 誠也, 高田 誠二1976 年 12 巻 5 号 p. 548-553
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーThe thermal emissivity ε of a freely radiating body with a partially transparent surface layer and the dependency of ε on the thickness d of the layer are studied. Cholesteric liquid crystal films of various thicknesses (d=13∼33μm) and a sufficiently thick (d≅1mm) one, prepared on an Al evaporated, reflecting glass surface, were chosen for experiment.
When examined by a pyroelectric IR sensor, a relationship which includes a differential coefficient may be established between the radiance and the temperature of the film. The differential coefficient may be compared with that of blackbody radiation. The comparison has given the value of ε as a function of d, and the absorption coefficient to be 7.24×10-3μm-1.
The dependency of ε on d thus obtained is consistent with the formulation developed here after McMahon's theory of radiant characteristics of the partially transparent films.抄録全体を表示PDF形式でダウンロード (896K) -
渡辺 紀之, 小宮 勤一, 長塩 圭一, 勝又 照久1976 年 12 巻 5 号 p. 554-559
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーCritical flow nozzles are commonly used as gas flow measuring instruments. However, this device has not received the degree of users acceptance it deserves, probably primarily because of the lack of a convenieht source of design data.
This paper presents essential datum for designing a small critical flow nozzle to be used as a standard of gas flow rate. The relation between the nozzle dimensions such as the entrance radius, the throat length, the exit diffuser angle, and the discharge coefficient is experimentally studied and it is found that the ratio of throat diameter to its length gives a great effect to discharge coefficient and that the other factors do not count so much.
From this result, a geometrically similar nozzle is designed and a certain relation between the Reynolds number and the discharge coefficient is obtained. The conclusion is that the discharge coefficient can be inferred by utilizing this relation if the nozzle is geometrically similarly made.抄録全体を表示PDF形式でダウンロード (872K) -
理論(第1近似)西山 卓, 増尾 龍一, 前田 親良, 村岡 茂信1976 年 12 巻 5 号 p. 560-564
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーDeformation of a single flat spring in a fulcrum is analyzed. Two fundamental differential equations are solved using the first order approximation in the angle θ between a tangent at an arbitrary point P on the spring and a vertical line.
It is found that the problems on the deformation of the single flat spring can be evaluated algebraically using only the addition theorem for hyperbolic functions, and that the deformation curve of the flat spring is expressed by a part of hyperbolic or exponential curve.
Application of the fulcrum of a single flat spring to a scale is presented in comparison with the knife-edge fulcrum.抄録全体を表示PDF形式でダウンロード (1842K) -
高速域での素子の挙動木下 修, 大島 康次郎1976 年 12 巻 5 号 p. 565-570
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーThis paper presents the experimental data concerned with the influences, of cavitation on the characteristics of the fluidic element. With reference to each degree of cavitation as defined by a cavitation number K, the variations of the control port pressure Hco, the vortex center pressure Hvor, the output port pressure Po and the attachment length La are investigated.
The dimensions of the fluidic element used are: main nozzle width bs: 4mm (or 3mm), aspect ratio: 2.0 (or 2.7), offset D: (0-3.0)×bs and wall angle α: 15°. Using water of 21∼22°C as the operating fluid, the studies were performed under the conditions of main jet velocity vs=4.0∼33.0m/s and various offsets.
The following experimental results are obtained:
(1) Cavitation appears and develops in the both shear layers of the jet. It appears first on the attached side and later on the unattached side. When the Reynolds number is increased further, the both leading locations of cavitation inception move upstream.
(2) Hco and Hvor hold linearity with vs2/2g and La is also constant until cavitation develops widely and swallowed by the vortex.
(3) The ratio of the output pressure to supply pressure does not reduce so soon as cavitation appears, but falls in the higher speed region because of the increase of cavitation loss.抄録全体を表示PDF形式でダウンロード (2019K) -
仁田 工吉, 沖津 泰, 鈴木 茂行1976 年 12 巻 5 号 p. 571-574
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーWith the rapid progress of power electronics, the thyristor chopper technique has been successfully used in the speed control of a d.c. series motor. In such a speed control system, the thyristor chopper is generally connected in series with the armature or the field winding, and with this series connection the motor can not be operated at a mode of regenerative braking.
In this paper, we consider a new speed control system which has the thyristor chopper connected in parallel with the field winding and also has an auxilliary diode bridge circuit in order to rectify the field current. In this system, the motor, is controlled by only one thyristor chopper at the power running mode as well as at the regenerative braking mode.
We analyze the speed-torque characteristics of the above mentioned system and discuss the dynamic performances theoretically and experimentally.抄録全体を表示PDF形式でダウンロード (529K) -
篠宮 栄, 岡田 昌丈1976 年 12 巻 5 号 p. 575-580
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーIt is well known that various performance characteristics of a 2-phase servomotor can be investigated by using its equivalent circuit parameters.
In this paper, a method to determine the equivalent circuit parameters (r1, xm, r2, x1) of 2-phase servomotor is described.
In this method, (1) r1 is determined by measuring the d-c resistance of the reference windings, (2) xm, r2 are determined by measuring both the maximum braking torque [Tod]max and the revolving velocity correspond to [Tod]max, on the d-c braking torque characteristics of 2-phase servomotor, (3) x1 is determined by measuring the applied voltage, input current and the tuning current of the reference windings at zero speed.
To check the accuracy of this proposed method, the average torque and the alternating torque characteristics of the sampled 2-phase servomotor are calculated by using the equivalent circuit parameters which is determined by means of this method, and these calculated results are compared with the experimental ones.
The agreement is satisfactory, and this method was found useful.抄録全体を表示PDF形式でダウンロード (727K) -
藤堂 勇雄1976 年 12 巻 5 号 p. 581-587
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーTransfer functions for 1-2 pass heat exchangers are obtained in a simplified form relevant to computer programming. In the transfer functions, the distributed temperature responses in the axial direction are related to changesin flow rates of both tube-side fluid and shellside fluid. A good agreement is obtained between the theoretical frequency responses and experimental data.
The relations between the outlet temperature responses and the distributed ones are demonstrated both in the frequency-and the time-domains: (1) The frequency transfer functions are illustrated on a complex plane as a function of circular frequency and distance. (2) Using Johnson's method, the time responses of fluid temperature profiles to a step change in flow rate are calculated from the transfer functions.
It is shown that the distributed temperature response gives much information on the dynamic behavior of heat exchangers.抄録全体を表示PDF形式でダウンロード (881K) -
戸苅 吉孝, 高松 武一郎1976 年 12 巻 5 号 p. 588-592
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーIn computer control systems, especially in a DDC system, digital controllers act according to the computer control algorithms.
In this paper the discrete PI control algorithm in DDC systems is discussed. Using the method by J.W. Sten to obtain the two time constants and dead time for a second-order process, the general solutions that give a set of optimall control parameters are obtained. In order to find the optimal control parameters the conjugate gradient method is used.抄録全体を表示PDF形式でダウンロード (664K) -
樋口 龍雄, 古山 健造1976 年 12 巻 5 号 p. 593-599
発行日: 1976/10/30
公開日: 2009/03/27
ジャーナル フリーTernary fail-safe logic systems have an important advantage in that the systems can clearly distinguish a failure state corresponding to a truth value φ from normal signals corresponding to truth values 1 and 0. For practical fail-safe systems both safety and reliability are required. The magnetic core itself has high reliability and is suited for a ternary fail-safe logic element.
This paper presents a practical realization of ternary Φ type fail-safe logic circuits using ferrite cores. Various functions such as NOT, OR, AND, Delay, NOR and NAND which are fail-safe can easily be obtained by using ferrite cores. The equivalent circuit in a magnetic circuit may be drawn from the measured values on the reciprocal of switching time versus the applied pulse amplitude which is well known as a basic characteristic of a core. On the basis of the equivalent circuit obtained thus, operations of fail-safe logic circuits can be made clear. A practical realization of the failsafe logic function X+Y is shown as an example.抄録全体を表示PDF形式でダウンロード (1969K)
- |<
- <
- 1
- >
- >|