巻号一覧
25 巻, 10 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
- >
- >|
-
Fumio TANAKA, David P. DEWITT1989 年 25 巻 10 号 p. 1031-1037
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーA new radiation thermometry method uses a reciprocity relationship between two uniquely defined spectral emissivities of a target whose surface characteristics may change during the period of observation. Emissivity variation of galvannealed steel due to iron-zinc alloy crystallization was measured experimentally, and the relationship between two emissivities at 1.6 and 3.0μm was defined over the temperature range 700 to 800K. Temperatures calculated by the new method showed excellent agreement with the true temperatures (±2.7K) while the conventional spectral and ratio methods yielded agreement within ±33 and ±4.4K, respectively. This method has potential for a wide variety of applications due to the simplicity of its principle.抄録全体を表示PDF形式でダウンロード (843K) -
花泉 弘, 藤村 貞夫, 豊田 弘道1989 年 25 巻 10 号 p. 1038-1045
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーEffects of profile of spectral emissivity are analyzed on temperature estimation in the wide band radiation thermometry. These effects are based on the fact that average emissivity used for temperature estimation is a weighted mean of the spectral emissivity through the observation band of a radiometer, and that the weight is described by spectral radiance of a blackbody. The results of the analysis are as follows:
(1) If an object is not a graybody, the mean emissivity depends on the temperature of the object,
(2) Emissivity measured by changing the temperature does not correspond to the mean emissivity, because the weight of the measured emissivity is the differential of the spectral radiance of the blackbody with respect to the temperature, and
(3) The difference between the mean emissivity and the measured one causes the biased error in the temperature estimation, where, amplitude and sign of the error depend on the shape of the spectral emissivity in the observation band.
These effects are quantitatively evaluated by numerical simulation, and qualitatively confirmed by actual measurement of temperature of water vapor and of a silicon plate.抄録全体を表示PDF形式でダウンロード (942K) -
堀口 和己, 西村 卓也, 長田 朗, 富田 英幸1989 年 25 巻 10 号 p. 1046-1053
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーDead-beat control is the one that can drive the output to the reference input in finite steps. But, it has two defects: 1) transient responses of the system, that is, the output response before the settling time and the control input response are not considered, 2) the system is not robust.
In this paper, we propose a design procedure for dead-beat control systems compensating these defects, based on two-degrees-of-freedom compensation scheme. We aim to design the control system satisfying the following two specifications.
a) For reference input of step function, the output is settled in finite steps, then the output error before the settling time and the control input are controlled moderately.
b) For disturbance of step function, the output is settled to zero in finite steps, then the output error before the settling time and the control input are controlled moderately.
Due to a), transient responses of the system are considered, and due to b), robustness of the system is improved.
The problem to design the control system specified above is reduced to linear optimization problems with respect to the parameters presented at the cost of increasing the settling time, and the solutions can be obtained with ease.抄録全体を表示PDF形式でダウンロード (830K) -
古谷 雅年, 大森 浩充, 佐野 昭1989 年 25 巻 10 号 p. 1054-1061
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーIn the least squares (LS) type of parameter estimation, we are often encountered with ill-posed cases where an input covariance matrix becomes nearly singular due to use of band-limited input signals. In order to attain stabilized convergence of the estimation, the present paper proposes a generalized regularization scheme using multiple regularization parameters and an a priori estimate, and gives analytical discussions on their optimal values which minimize the mean square error (MSE) or the estimated MSE using only accessible signal data. It will also clarify the relationship with the optimal number of truncated smaller eigenvalues in the eigenvalue decomposition. These results discover successfully an explicit form of the optimal initial condition for the recursive least squares algorithm so that the MSE can be minimized at each instant. Finally an on-line regularization scheme is presented.抄録全体を表示PDF形式でダウンロード (1076K) -
須田 信英, 万 柏平, 上野 一郎1989 年 25 巻 10 号 p. 1062-1068
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーThe orders of infinite zeros, defined as the orders of zeros of G(1/λ) at λ=0, where G(s) is the given system transfer function matrix, are important parameters in the linear system analysis.
In this paper, we investigate the generic orders of infinite zeros for multivariable structured systems. By a structured system we mean a linear, time invariant system, whose coefficient matrices contain some fixed zero elements. It is assumed that the elements other than fixed zeros may vary independenty of one another and take any real value except zero. The generic orders are orders for almost every possible set of values of the variable elements.
The signal flow graph and a modification of the Mason's formula have been used. It is observed that the generic orders of infinite zeros are obtained from the minimum number of integrators included in a separate set of forward paths between k pairs of inputs and outputs, k=1, 2, …, σ, where σ is the rank of the system transfer function. Furthermore it is possible to select such pairs with nesting property, that is, the (k-1) pairs used to obtain first smallest (k-1) orders are included in k pairs for obtaining the k-th smallest order. Finally a sufficient condition is derived for the order of infinite zeros to be identical for every possible set of values of variable elements.抄録全体を表示PDF形式でダウンロード (901K) -
山本 透, 大松 繁, 石原 弘一1989 年 25 巻 10 号 p. 1069-1075
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーPID control algorithm has been used for several process control systems up to now. However, it is difficult to determine appropriate gains in this control algorithm. For example, we may have excessive overshoot and long settling-time as adjusting gains. Therefore, it is important to determine and adjust PID control gains since control performance depends on these gains. Takahashi has proposed a method to determine PID gains by using characteristics of a step response. However, we may not be able to get a control performance the case of time-variant process control system.
In this paper, in order to overcome these problems we construct a self-tuning PID control system based on the generalized minimumvariance control. We derive a system model from the PID control scheme. By the least squares identification method, the unknown parameters included in the system model are estimated. Using the estimated parameters we obtain prediction values of the system. Letting the prediction values equal a desired signal of the process, we construct a selftuning PID control system. To show the effectiveness of the present self-tuning control system some numerical simulation results are illustrated. Finally, the proposed control method is applied to the pressure control problem for the pressure tank and the quantitative effectiveness is shown.抄録全体を表示PDF形式でダウンロード (805K) -
崔 然旭, 荒木 光彦1989 年 25 巻 10 号 p. 1076-1082
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーIn industrial multivariable plants, it is often the case that the plant outputs are detected not simultaneously but serially. In this paper, the problem of estimating the state vector of the plant based on the data obtained from such a detecting scheme is considered, and a special type of observers (referred to as “serial-sampling” type observers) which renew their internal states when-ever a new data is obtained is proposed. It is proved that such observers can be constructed for almost all sampling periods if the plant is observable as a continuous-time multivariable systems. The results of simulation indicate that serial-sampling type observers can estimate the state of the plant more quickly than the ordinary observers.抄録全体を表示PDF形式でダウンロード (847K) -
河村 仁, 土谷 武士1989 年 25 巻 10 号 p. 1083-1090
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーOne of the useful method to improve a designed control system performance is utilization of the future desired signal and/or disturbance signal, if possible. Until now, optimal preview control system and preview pre-compensator are considered on the basis of optimal regulator theory and model matching approach. Effectiveness of the preview control system is shown in the application to servo system and robot manipulator control etc. Usually, preview control system has excellent phase characteristics compared with the usual control system which does not utilize future desired signal. In order to make clear the characteristics of the optimal preview control system, asymptotic characteristics of the preview control system with respect to near-zero weighting factor for input variable term in the quadratic performance index. In this paper, the asymptotic characteristics for the following three points are shown (1) closed loop poles (2) pulse transfer function (3) performance index value.
Pulse transfer function from desired signal to controlled variabvle approaches to unity and that from disturbance signal to controlled variable approaches to zero by utilizing future desired signals. Performance index value approaches to zero on approaching zero weighting factor for input variable in the quadratic performance index. Above considerations are carried out for two cases control system with and without input delay time due to processing time for control algorithm.抄録全体を表示PDF形式でダウンロード (964K) -
中村 達也1989 年 25 巻 10 号 p. 1091-1097
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーIt is necessary to compute binding forces in dynamic simulation of a linked mechanism interacting with an environment, since the constraints change frequently. An efficient solution method is also required especially for the mechanism with many degrees of freedom. This paper deals with a legged robot. Equations of motion and constraints are formulated and their efficient solution method is proposed. Then it is applied to the three dimensional dynamic simulation of the locomotion of a quadpod robot as an example.
The method is executed in the following two steps: The first computes link accelerations and joint binding forces as functions of the binding forces of the first joint for each leg independently, using leg equations. The second determines body accelerations and the binding forces of the first joints using body related equations. Since the entire process is divided into several subprocess, some of them are executable parallel and the dimension of equations solved at a time becomes small.
Two schemes of simulation are also proposed. One specifies the motion of a robot by the joint angles. The other does by the joint torques. This paper deals mainly with the former and discusses a statically indeterminate problem concerning the reaction forces from the ground. An elastic model of the ground with friction is proposed to resolve the problem. Since friction makes equations nonlinear, an iteration method is used to solve it where slippage is detected by aproximation. As for the latter, it is pointed out that slippage and collision problems are unavoidable and left as a future work.
The proposed methods are general enough for the application to a linked mechanism besides a legged robot.抄録全体を表示PDF形式でダウンロード (1024K) -
小山 紀, 天沼 稔, 木村 靖志, 原田 正一1989 年 25 巻 10 号 p. 1098-1103
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーRecently, the solenoid valves are often used to positioning the air cylinder to the desired position. The positioning accuracy of cylinder is effected by the accuracy of total flow volume through the valve. But the solenoid valve is switching element essentially then it can not control the air to very small flow rate in comparison with such as proportional valves. It is very important to improve the performance when the small flow volume is needed for the high accuracy positioning, so the method by using serial solenoid valves is proposed in this paper. That is composed by double serial stages of solenoid valves and chambers inter the valves. When the valves open without overlap, the air through the valves is pooled in the chamber once. If the chamber volume is set small, the flow volume can be limited to small naturally.
The step feed of air cylinder with very small displacement is executed by limiting the charging and discharging air in this manner. One of the serial valves is used to the charging side and the other is used to discharging side of cylinder. As the results of step feed, the stepping displacement is effected not only supply pressure and chamber volume also the initial pressure in the cylinder, and the stepping displacement can be very small when the initial pressure is set to supply pressure. The positioning against the desired position is carried out by constructing the feed back loop using micro-processor unit. The cylinder moves to near the desired position quickly by opening the all valves at the same time. Then moves slightly with step feed toward the desired position. The total accuracy of experimental positioning is better than that of conventional valve method and it is enable to execute the positioning within 0.004mm from the desired position. But the settling time takes fairly long and it must be improved to practical use.抄録全体を表示PDF形式でダウンロード (822K) -
井上 正夫, 二宮 敬虔1989 年 25 巻 10 号 p. 1104-1110
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーIntroducing magnetic suspension to the momentum wheel of an artificial satellite is expected to enhance the reliability and the maneuvability of the satellite. However, as the stiffness of the magnetic bearing is rather small, the attitude motion of the satellite may be coupled with the wheel precession to become unstable. Although the attitude motion may be able to be stabilized at the cost of reducing the stability of the wheel precession when the tilting angles of the wheel are controlled actively, it has been difficult to get sufficient stability margin both in the attitude motion of the satellite and in the precessional motion of the wheel without attitude sensors. This paper proposes the new strategy to offer large damping factor both to the satellite motion and to the wheel precession without attitude sensors. That is achieved by introducing an observer which estimates the attitude-rate of the satellite. It is provided that the observer-based compensation is equivalent to the low-pass-filtering whose cut-off frequencies are as low as those of the attitude motion of the satellite. Some numerical examples displayed that the system has good performances.抄録全体を表示PDF形式でダウンロード (826K) -
前田 太郎, 舘 〓1989 年 25 巻 10 号 p. 1111-1118
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーOne of the distinguished characteristics of human binocular visual space is the phenomenon which is represented as horopter and parallel or distance alley. In this paper, neural network models of space perception using binocular vision are presented to find out how the bipolar parallax and the bipolar latitude are mapped onto the depth sensation. One of the proposed models based on the neural network model shows the characteristics of horopter and alley which is similar to the human space perception. A structure of the model represents how the space perception is generated in the human brain.抄録全体を表示PDF形式でダウンロード (1067K) -
線型フィルタの階層的適応手法高橋 弘太, 山崎 弘郎1989 年 25 巻 10 号 p. 1119-1125
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーAn intelligent sound sensing system is proposed. The system can sense a target signal with better signal-to-noise ratio automatically. The target sound is picked up with disturbing ambient noise sounds by multiple microphones. The multiple outputs of microphones are processed by an adaptive linear filter.
In this paper, a new learning algorithm of an adaptive filter is described. For adaptation, this method needs only a simple cue signal. The cue signal is used for identification of the target signal. Any signal which correlate with the power level of the target signal can be used as the cue signal. As the results, it is not necessary to know transfer functions between sound sources and microphones and/or signal waveforms of sound sources for training. Therefore, spatial arrangement of multiple microphones and relative locations of the target signal source and noise sources can be chosen arbitrarily.
The efficiency of this method is demonstrated by numerical simulations of the self-adapting multiple microphone system.抄録全体を表示PDF形式でダウンロード (968K) -
野本 弘平, 近藤 倫正1989 年 25 巻 10 号 p. 1126-1133
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーFuzzy reasoning is able to utilize human experimental rules for driving numerical results. So, recently, some automatic control systems based on fuzzy reasoning have been developed. But the conventional fuzzy reasoning which is used in control has difficulties for application to parameter estimation. Because it needs inputs which are always meaningful and doesn't have property like convergence.
In this paper, we propose recursive fuzzy reasoning. This reasoning inputs phenomena induced by a parameter and estimates the value of the parameter, gradually. This algorithm has a learning membership function which narrows range of the parameter.
We have applied the recursive fuzzy reasoning to a PI autotuning controller. Usual controlled systems are so complicated that a skilled human operator's technique is often more desiable than that based on mathematical models. Besides, the auto-tuning is a kind of parameter estimation. So the recursive fuzzy reasoning is suitable for this problem. Furthermore, we use other two new techniques for this application. They are the evaluation of unsatisfaction and the extraction of characteristic parameters.
Simulation tests are presented. The recursive fuzzy reasoning automatically tunes control parameters during ordinaly on-line control operation without any special test signal.抄録全体を表示PDF形式でダウンロード (1053K) -
西村 金太郎, 住本 哲宏1989 年 25 巻 10 号 p. 1134-1136
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーThe fire prevention detector that used the plastic fiber was investigated. Abnormal air temperature in the house affected the optical output intensity of the plastic fiber. Therefore, the optical output intensity of the fiber was detected, and the warning of the abnormal air temperature could be given with the reformative equipment.抄録全体を表示PDF形式でダウンロード (924K) -
師玉 康成, 山浦 弘夫, 太田 豊生1989 年 25 巻 10 号 p. 1137-1139
発行日: 1989/10/30
公開日: 2009/03/27
ジャーナル フリーA new first-order algorithm is adopted to gain the solution of nonlinear optimal control with terminal constraints. Together with the fixed point theorem, variational principle is applied to settle the terminal point.
This method has the advantage that the choice of first approximation in the variational methods is far less sensitive compared with the traditional ones.抄録全体を表示PDF形式でダウンロード (413K)
- |<
- <
- 1
- >
- >|