巻号一覧
20 巻, 12 号
選択された号の論文の16件中1～16を表示しています
- |<
- <
- 1
- >
- >|
-
村田 弘志, 相良 節夫1984 年 20 巻 12 号 p. 1073-1080
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーIn the design method of the deadbeat digital controller, the Z domain approach using the pulse transfer function of the process or the time domain approach using state vector equation is generally used.
In this paper, we propose a time domain design for an optimal deadbeat controller which can minimize the quadratic performance index for the manipulated variable and deviation or controlled variable.
The sequence of the optimal control signal which makes the deadbeat response for step, ramp, or constant acceleration variation of the desired value and step-wise disturbance can be easily obtained from the computation of the matrix constructed by the impulse response sequence of the system.
Then, an optimal deadbeat controller is designed by using sequence of the control signal.
The method proposed here is very simple comparing with other methods and can be systematically applied to the system including the lag time or the restriction of the manipulated variable.抄録全体を表示PDF形式でダウンロード (1125K) -
井前 讓, 箱守 京次郎1984 年 20 巻 12 号 p. 1081-1088
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーThis paper presents a strong variation algorithm for solving optimal control problems whose dynamics are described by ordinary differential equations with discontinuous right-hand sides.
It is proved that the accumulation points of the sequence of controls generated by the algorithm, if exist, satisfy maximum principle type necessary conditions for optimality. The major difficulties which arise from the existence of “separating hyperplanes” dividing the state space into subregions are that state trajectories do not necessarily exist on the whole time interval, and that adjoint variables may have the jump-discontinuities at hyperplane passing times. Both of the existence of state trajectories and jump-phenomena of adjoint variables are discussed in detail. Moreover, two examples with separating hyperplanes are shown to illustrate the effectiveness of the present algorithm.抄録全体を表示PDF形式でダウンロード (919K) -
中溝 高好, 小林 伸明1984 年 20 巻 12 号 p. 1089-1094
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーThe problem to be considered in this paper is to find general duality between a control problem and a state estimation in linear time invariant systems.
A dual form of the filter equation, derived previously by present authors for a linear stochastic system with completely unknown disturbances, is introduced. Then it is shown that the above dual equation yields the solution of an output zeroing problem.
On the other hand, a limiting form can be derived from a standard linear regulator problem by letting the weighting matrix associated with the state vector tend to infinity. Futhermore, it is poited out that the limiting form also resolves itself into the solution to the output zeroing problem. From this observation, it can be shown that the above limiting form of standard linear regulator coincides with the dual form of the filter equation with completely unknown diturbances. Coincidentally, the solution of the output zeroing problem which minimizes the input cost function is explicitly obtained.抄録全体を表示PDF形式でダウンロード (647K) -
水野 毅, 樋口 俊郎1984 年 20 巻 12 号 p. 1095-1101
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーAn unbalance on a rotor causes vibration phenomena in the rotating machine. Thus balancing is an important process in manufacturing rotary machinery. Since perfect balancing is rather difficult there are often residual unbalances on the rotor. Moreover the rotor sometimes becomes unbalanced while the machine is in operation. To overcome these problems a new control method is presented for totally active magnetic bearing systems. By this method an unbalanced rotor can be suspended without unbalance vibrations by the magnetic bearing. The method is based on the theory of the output regulation with internal stability. The ways of designing the compensator are described as follows.
(1) The system is stabilized by state feedbacks.
(2) Since unbalance forces cannot be measured directly, an observer is constructed in order to estimate unbalance forces from measurable quantities: displacements and velocities of the rotor.
(3) Using the outputs of the observer the unbalance forces are cancelled by controlling electromagnetic forces of the magnetic bearing. It is shown by numerical examples that the designed system can suspend an unbalanced rotor without whirling motions which appear when the compensation is not done. Moreover the effectiveness of the compensation for unbalance is experimentally studied by an instrument with Gyro mechanism.抄録全体を表示PDF形式でダウンロード (1402K) -
市川 邦彦1984 年 20 巻 12 号 p. 1102-1106
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーIt is well known that any controllable finite dimensional plant, whether it is stable or not, can always be stabilized. Smith's method has long been regarded as useful for the control of the plant with delay. As is known, Smith's method is valid only for the stable plant; the reason is explained in this paper from the view point of unstable pole zero cancellation. Instead, the control scheme based on the calculated value of x(t+L) using the parameter values of the plant achieves finite pole assignment regardless of the plant stability.
The frequency domain pole assignment technique for the finite dimensional plant is applied to the control of the unstable plant with delay in this paper Quite the same result as finite pole assignment could not obtained, but it is demonstrated that the unstable plant with delay cannot always be stabilized according to the relation between the magnitude of unstable poles and delay time. Stability region depicted in the gain-delay plane is presented for the case when there exists only one unstable pole. Simulation result shows that the unstable plant with delay is controlled with satisfactory stability by the proposed design method.抄録全体を表示PDF形式でダウンロード (707K) -
岩井 善太, 井上 昭, 石飛 光章, 永田 正伸1984 年 20 巻 12 号 p. 1107-1113
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーThis paper proposes a design method of a single-input single-output model reference adaptive control system with desired rate of exponential convergence. The design scheme is based on the concept of an adjusting law used in an adaptive observer with an exponential rate of convergence developed by authors. State variable filters are used in the estimation procedure of the output error between the plant and the model. Additional filters are also included in the adaptive scheme in order to avoid the use of derivatives of output signals. The control input signal is determined by equations making the estimated output error go to zero with an exponential rate. In addition, the augmented error vanishes exponentially. Therefore, the output error decreases with an exponential rate of convergence. The convergence rate is specified arbitrarily.
The effect of the design scheme is confirmed by applying it to a second order thermal experimental system.抄録全体を表示PDF形式でダウンロード (944K) -
河野 克己, 小坂 満隆, 宮本 捷二1984 年 20 巻 12 号 p. 1114-1121
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーGenerally, various approaches, like optimization, stabilization, adaptation and so on, can be taken to solve a control problem. Thus, there exist several kinds of algorithms to solve a certain problem and each algorithm's characteristics depend on the approach to be taken. In the case that these algorithms are applied to practical system problems, it is necessary to examine the algorithms' effectiveness and select a suitable one among usable algorithms.
On the other hand, in a man-machine system, a human operator occasionally selects the better one by use of his experimental know-how or a priori information. But it is very difficult for the operator to make a good decision continuously in a real time system.
In this paper, an on-line algorithm selection problem is considered. Firstly, the algorithm selection process is formulated as a fuzzy decision-making problem and a fuzzy algorithm selection method is proposed. If-then type rules are used to reason the effectiveness of the algorithms. A priori information and know-how is utilized in the reasoning rules. Secondly, this method is applied to the guidance and control problem in a command and control system. Finally, simulation is conducted to evaluate the effectiveness of the proposed method.抄録全体を表示PDF形式でダウンロード (1079K) -
山口 隆美, 吉川 昭, K.H. PARKER1984 年 20 巻 12 号 p. 1122-1129
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーTwo features should be kept in mind when the parametric spectrum analysis, such as AR or MEM method, is applied to the turbulence in the aorta. Firstly, the broad frequency band width spectrum is the main interest in turbulence research. Spectra of this kind are apt to be regarded as originated from noises rather than a signal of interest in other fields. Secondly, in an unsteady stochastic process, only ensemble average technique is strictly valid for the definition of statistical measures. Spectral calculations, therefore, should be based on an ensemble average, while parametric methods so far reported are not.
In the present study, a method of spectral calculation based on the ensemble average technique, which is a natural expansion of the Burg's maximum entropy method, is proposed. This method is applied to a set of numerical simulation data which simulates turbulent velocity fluctuations data measured in the canine aorta with a hot-film anemometer. Autoregressive order of the spectral calculation is chosen by the minimum AIC (Akaike's Information Criterion) method. It is shown that the proposed method has a very good performance in the short time quasi-steady spectral analysis of the turbulence in the aorta.抄録全体を表示PDF形式でダウンロード (1077K) -
武冨 喜八郎1984 年 20 巻 12 号 p. 1130-1134
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーThe new measuring method of a complex dielectric constant which can observe the wide-range frequency characteristics of materials, has been developed.
In this paper, the theory of measurement, its results and discussions are described.
In this method, a simple measuring circuit is employed, and it fits particularly in the measurement of liquid mateials.
While observing the spectrum on CRT, one can easily recognize the wide-range frequency dependence of dielectric characteristics.
And also, by adjusting a plunger in specimen waveguide, even the small dielectric losses of the dielectric constants of materials can be measured.抄録全体を表示PDF形式でダウンロード (1082K) -
とくに手本パターンの発生法を中心に原 靖彦, 浜田 利満, 岡本 啓一, 中川 清, 江口 一彦1984 年 20 巻 12 号 p. 1135-1142
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーFor inspecting LSI reticle patterns automatically with high accuracy, pattern inspection method by comparing reticle patterns with its design data is described.
This method utilizes the data, which controlls a pattern generator for drawing reticle patterns, as a design data, The data consist of an aggregate of rectangles which form the reticle patterns. The data are stored in a magnetic tape in the form of rectangle parameters (center position, width, height, slant angle).
In order to compare the optically detected reticle patterns with the design data, the data are converted to a video signal along the detector's raster scanning line of the patterns. The method for the conversion and block diagrams of the conversion circuits are presented. The circuits generate a video signal from the design data, at high speed. Inspection is then performed by comparing the generated signal with the detected reticle pattern signal.
The structure of the reticle inspection system utilized the developed method is briefly described. The system detect defects as small as 4μm. The time of inspection is 45 minutes for 90×90mm reticle.抄録全体を表示PDF形式でダウンロード (1939K) -
森内 勉, 柏木 濶1984 年 20 巻 12 号 p. 1143-1149
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーThe upper bound of crosscorrelation of M-sequences which have the same period but different characteristic polynomials is shown theoretically.
An M-sequence generated by an n-th degree characteristic polynomial which is a primitive polynomial over GF(2) is constructed, in case of the period N=2n-1 is nonprime, by the combinations of s nonmaximum length sequences with length q generated by an n-th degree irreducible polynomial of exponent q, where q=N/s and s is a positive integer and q>s. Therefore the primitive polynomials can be classified into the same class, if the nonmaximum length sequences are the same under the sampling of every s bits from the original M-sequences.
Using this classification, it is shown that the crosscorrelation between the M-sequences in the same class is larger than that of other M-sequences. The crosscorrelation function is obtained by the weights of the nonmaximum length sequences. And the maximum crosscorrelation value appears at the delay time of an integral multiple of q, when the two M-sequences are relatively phased so as for the initialpoints characteristic M-sequences to coincide, and then the crosscorrelation function is calculated with the delay taken from that phase.
The upper bounds of the crosscorrelation function of different M-sequences in case of s=5 and 7 are obtained.抄録全体を表示PDF形式でダウンロード (870K) -
西田 茂人, 中村 政俊, 柴崎 浩1984 年 20 巻 12 号 p. 1150-1155
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーSince the electroencephalogram (EEG) is used as a supporting method in the diagnosis of cerebral damage in a medical field, it is important to grasp the features of the EEG. At present, doctors read the EEG in the time domain qualitatively.
In this paper, we propose a method for optimally approximating power spectrum with distribution function. In this method, we assume a distribution function which is similar in form to the power spectrum of the EEG, and calculate the parameters of this distribution function by minimizing the square error of the area between the power spectrum of the EEG and the distribution function using the extremum search method. Then the features of the power spectrum of the EEG can be expressed by the parameters of the distribution function. In order to show the effectiveness of this method, using the parameters of the distribution function we calculate the feature parameters of the EEG which are closely related to the findings of doctors, and we judge the EEG quantitatively using these parameters. The results by this method agree approximately with the judgements of doctor.抄録全体を表示PDF形式でダウンロード (889K) -
佐藤 勝俊, 畑中 浩1984 年 20 巻 12 号 p. 1156-1160
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーThis paper presents experimental results of the frequency response of laminar proportional amplifiers (LPA). The LPA are manufactured by a wire electrical discharge machine instead of the photochemical etching process and so the accuracy of manufacturing the LPA is fairly good.
It is observed that pressure oscillation occurs at the blocked output ports with increasing the supply flow rate. The operating range of the LPA, therefore, is restricted by this oscillation. It seems that the oscillation is due to the edgetone occurring at the splitter of the output ports, but the frequency of oscillation observed is about a tenth of the wellknown experimental results of edgetone reported by G.B. Brown. It is found that the supply nozzle width should be wide and the aspect ratio be small to get a wide operating range.
The dynamic characteristics of the LPA depend strongly on the supply flow rate or the supply pressure. As the supply flow is increased, the pressure gain and the bandwidth increase with a remarkable resonant peak and the phase shift decreases.
It is also found from a low order lumped parameter analysis that the frequency response of LPA is well represented by the third order system with a transport time delay equal to twice an average particle transit time.抄録全体を表示PDF形式でダウンロード (622K) -
浅田 春比古, ヨセフ＝トーミ カマル1984 年 20 巻 12 号 p. 1161-1168
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーThe direct-drive arm that has no gears between motors and their loads has several important advantages including no backlash, small friction and high mechanical stiffness. The arm mechanism, however, becomes extremely massive, when each motor is directly attached to its jont along a serial linkage mechanism. The complicated dynamics resulting from varying inertia, interactions, and nonlinearities, is also more prominent than that of a robot with gears.
This paper describes a lightweight arm mechanism with invariant and decoupled inertia characteristics. Instead of having motors at serial joints, a parallel drive mechanism with a closed-loop five bar linkage is utilized. The dynamic behavior of this mechanism is analyzed and the condition for the elimination of the interactions and nonlinearities in the mass properties is derived. The decoupled and invariant arm dynamics significantly reduces the complexity of controlling the directdrive arm. In the latter half of the paper, a prototype robot developed on this basis is described. By using high torque brushless motors which were specially designed for the direct-drive robot, top speed and maximum acceleration were increased by an order-of-magnitude to about 10m/s and 5G, respectively.抄録全体を表示PDF形式でダウンロード (1634K) -
示村 悦二郎, 藤田 政之1984 年 20 巻 12 号 p. 1169-1171
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーPDF形式でダウンロード (359K) -
新中 新二1984 年 20 巻 12 号 p. 1172-1174
発行日: 1984/12/30
公開日: 2009/03/27
ジャーナル フリーPDF形式でダウンロード (346K)
- |<
- <
- 1
- >
- >|