All issues
Volume 5 (1969)
- Issue 5 Pages 401-
- Issue 4 Pages 295-
- Issue 3 Pages 201-
- Issue 2 Pages 103-
- Issue 1 Pages 1-
Volume 5, Issue 4
Displaying 1-13 of 13 articles from this issue
- |<
- <
- 1
- >
- >|
-
Kiyotaka SHIMIZU1969 Volume 5 Issue 4 Pages 295-304
Published: October 30, 1969
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe optimal satisfactory control of systems subject to external disturbance inputs is studied. The systems represented are essentially deterministic non-linear, continuous-type processes which are normally operating in the steady-state. The objective is on-line static optimization according to an economic performance criterion.
The optimal satisfactory control is formulated by combining concept of satisfactory control and concept of optimal control. In order to solve the optimal satisfactory control, several theories and techniques of decision making under uncertainties are developed to obtain necessary conditions for maximizing system performance. The Lagrange multiplier techniques for game solving under general inequality constraints are derived to calculate the formulated optimal satisfactory control.View full abstractDownload PDF (1245K) -
Takami KOYANAGI, Kazuo AIDA, Harushige UMEUCHI1969 Volume 5 Issue 4 Pages 305-312
Published: October 30, 1969
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA design consideration for a fast response radiation thickness gauge which uses a cylindrical ionization chamber as the detector is descrived. It is pointed out that, in realizing a fast response, the time lag in the response of the ionization chamber must be taken into account as well as that in the amplifier stage.
According to the expressions derived for the transient responses of the overall system, fundamental requirements are considered for attaining a fast response or a short rise time (0.1∼0.9) in the indicial response.
Besides the importance of using an amplifier with small time lag, it is shown that the ionization chamber has an optimum ratio of the diameter of the outer electrode to that of the inner electrode, depending upon the time lag in the amplifier. When the amplifier has a fairly large time lag, a small value of the ratio is preferable in the design of the ionization chamber. This will reduce the time required for positive ions to reach the outer electrode. On the contrary, in case that the time lag is relatively small, a large value of the ratio is recommendable so as to increase the output voltage due to electrons.
It is also shown that the response time can be improved more or less by either raising the voltage between the electrodes (V0) or lowering the filling gas pressure (p) or employing a filling gas with high mobility (μβ), but the these procedures are effective only within a limited range of the value of V0μβ/p.
The above considerations have proved to be in good agreement with the experimental results of the indicial and frequency responses.View full abstractDownload PDF (918K) -
Hiromitsu ISHII, Norio MUROI1969 Volume 5 Issue 4 Pages 313-319
Published: October 30, 1969
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn the case of discussing the properties of light sources and luminares, it is important to know the total luminous flux and the light distribution curve. We have studied on the automatic measurement of light distribution curve and total luminous flux for several years, and now, it is possible to measure within from a few minutes to scores seconds. In this paper, the authors described the general of the automatic light distribution measuring equipment and showed the integral method of luminous flux. In addition, we discussed the error of photometric system and made clear that the measurement by light distribution method is better than globe photometer.View full abstractDownload PDF (822K) -
Tadashi NAGATA1969 Volume 5 Issue 4 Pages 320-325
Published: October 30, 1969
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSProblems that are concerned with control and estimation of stochastic linear dynamic systems with noisy measurement subsystems have been investigated by many researchers, and important results have been obtained. That is, the duality of optimal control and optimal estimation, Fel'dbaum's dual control, and Meier's combined control etc. seem to be the representative fruits in this area. Proceeding with these studies, the following problems are being brought forward: the reliability of measurements, the decision of optimal control policy for the system with interrupted measurements, and the configuration of the optimal control system considering the cost of measurements.
In this paper, a measurement adaptive problem, in which the optimization is required for not only the plant but also the measurement subsystem, is discussed. First of all, it is shown that the optimization of the measurement subsystem with some constraints can be performed by off-line computation unlike the optimization of the usual control system, and a new method to solve this optimal measurement problem with the aid of measurement control matrix is given. Moreover, the concept of measurement sensitivity being proposed, the information value in the control system is evaluated on the basis of this concept. This measurement sensitivity is nothing but quantifying the value of information obtained by the measurement subsystem, and it is closely related with the optimization of the measurement subsystem.View full abstractDownload PDF (899K) -
An Approximate Method of State Estimation for Nonlinear Dynamical Systems with State-Dependent NoiseYoshifumi SUNAHARA, Koji YAMASHITA1969 Volume 5 Issue 4 Pages 326-334
Published: October 30, 1969
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, a method of stochastic linearization is demonstrated for the purpose of establishing an approximate approach to solve filtering problems of nonlinear stochastic systems with state-dependent noise in the Markovian framework.
The models of both the dynamical system and the observation process are described by nonlinear stochastic differential equations of Itô-type.
The principal line of attack is to expand the nonlinear drift term into a certain linear function with the coefficients which are determind under the minimal squared error criterion. Two methods of linearization are developed for the nonlinear diffusion term. The linearized models are thus characterized by the expansion coefficients dependent on both the state estimate and the error covariance.
A method is given for the simultaneous treatments of approximate structure of state estimator dynamics and of running evaluation of the error covariance, including quantitative aspects of sample path behaviors obtained by digital simulation studies.View full abstractDownload PDF (1040K) -
Hisashi OSAWA, Norihiko MORINAGA, Toshihiko NAMEKAWA1969 Volume 5 Issue 4 Pages 335-340
Published: October 30, 1969
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSAn analysis is presented on the output signal-to-noise ratios of the FM correlation system having an FM detector at each input channel of a conventional correlator, in which the input consists of a frequency-modulated signal combined additionally with stationary narrow-band Gaussian random noise.
First, a general expression is derived for the output signal-to-noise ratio. Next, as a practical case, a detailed calculation is made for the output signal-to-noise ratios when each input signal is the sinsoidal carrier frequency-modulated by a Gaussian random process and the integrating filter is of RC low-pass.
Some important characteristics are obtained as follows;
(i) For large input carrier-to-noise ratio the output signal-to-noise ratio has the constant value independent of the input carrier-to-noise ratio, but it is proportional to the product of the r.m.s. bandwidth of the input noise and the time constant of the RC low-pass filter, and the ratio of the r.m.s. bandwidth of the modulating signal to the r.m.s. bandwidth of the input noise.
(ii) For small input carrier-to-noise ratio the output signal-to-noise ratio is proportional to the square of the input carrier-to-noise ratio, the product of the r.m.s. bandwidth of the input noise and the time constant of the RC low-pass filter, and also to the square of the ratio of the r.m.s. modulation to the r.m.s. bandwidth of the input noise.View full abstractDownload PDF (656K) -
Hidekatsu TOKUMARU, Zenta IWAI, Kenji UEMOTO1969 Volume 5 Issue 4 Pages 341-347
Published: October 30, 1969
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe problem of noninteracting control in multivariable systems has been investigated by many authors and it has been known that the feedback control which realizes the noninteraction for linear multivariable systems is linear with respect to state-variables and includes a number of arbitrary parameters. Thus, one of the important design problems of noninteracting control systems is to determine the value of these arbitrary parameters.
In this paper, the method which determines the optimal value of these parameters in the sense of optimal tracking control is proposed. That is, for a given quadratic performance, these parameters are chosen so as to minimize the quadratic performance loss used in the theory of adaptive control and sensitivity analysis. In this case, the noninteracting control system becomes a suboptimal control system in the sense that the quadratic performance loss is minimum. The special case is also shown that under some conditions these arbitrary parameters can be determined so as to minimize a given quadratic performance itself. In such a case, the noninteracting control system is consistent with the optimal tracking control system.View full abstractDownload PDF (780K) -
Yoichi KAYA1969 Volume 5 Issue 4 Pages 348-357
Published: October 30, 1969
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSAlmost all papers dealing with modeling problem focus their attentions in how to find optimum values of parameters of a system model, of which form is given a priori. In practical situations, however, it is not known in advance whether the given form of the model is valid as a mathematical expression of the system behaviour, so it is necessary to test goodness of fit of the model.
A method to test goodness of fit of a transfer function model of a dynamical system is proposed in this paper. The hypothesis that the model is valid is rewritten into the hypothesis that
E[I]=0
Where I is a random variable, a function of the system input and output, and has following properties;
1) I is normally distributed, and
2) every sample of I is independent.
The hypothesis is then able to be tested by the well-known method of Student. Two examples of simuation tests, discussion on the data length required to detect the invalidity of the model, and the method of test when model parameters have dispersion are described.View full abstractDownload PDF (2066K) -
Iwao MORISHITA1969 Volume 5 Issue 4 Pages 358-367
Published: October 30, 1969
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper deals with a perceptron-type element in which variable weights change automatically following a certain rule of growth. An analysis of its dynamic behavior is described together with some simulation results.
The element is a summing device. Its output y(t) is a weighted sum of its inputs xi(t), i=1, 2, ……, N, i.e.,
y(t)=NΣi=1wi(t)xi(t)
The inputs xi(t) are assumed to be zero-mean signals, but not restricted to binary signals. Changes in the weights wi(t) are described by the differential equations
Tdwi(t)/dt+wi(t)=axi(t)sgn[y(t)] i=1, 2, ……, N.
A detailed investigation of the solutions of the above equation shows that the element has a strong tendency to separate its inputs into a family of principal components and to pick out the greatest component as its output. This property enables it to perform a variety of types of information processing such as factor analysis, signal filtering, pattern dichotomy, majority decision logic and memory.View full abstractDownload PDF (1133K) -
Katsuhiko FUJII, Kazuo NAKAO1969 Volume 5 Issue 4 Pages 368-377
Published: October 30, 1969
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper proposes a method to measure the dynamic characteristics of nonlinear systems whose topological structures are not known. The Volterra series model is introduced for this purpose. First of all, a certain measure which indicates the degree of the approximation of the Volterra series model to the system is defined. The order of the model is chosen beforehand on the basis of the value of this measure. In turn, the Volterra kernels which represent the dynamic characteristics of the system can be determined by evaluating the responses of both the model and the system to deterministic signals. By using these Volterra kernels a digital model of the system can be constructed.
By this method it is possible to measure dynamic characteristics of any nonlinear systems which can be described by continuous functionals. It is shown that the degree of approximation of this method can be specified arbitrarily as desired. It is also shown that the digital model is easily constructed when the system can be represented by Volterra functional series whose order is less than four. Several examples to which the proposed method is applied are presented.View full abstractDownload PDF (1225K) -
The Learning Control of a Linear System Containing Unknown Functions of Measurable Random ParametersSetsuzo TSUJI, Kousuke KUMAMARU1969 Volume 5 Issue 4 Pages 378-385
Published: October 30, 1969
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSMany control systems in the technical processes have various unknown elements, which are generally time-varying.
This is usually due to the change of environment around the control objects, arising from the time-variation of physical parameters (such as temperature, pressure, source-voltage, and so on).
Therefore, it is necessary to regard these elements as the unknown functions of such parameters, not time-varying, and to learn these functions on-line.
In some parts of control design problems, it is desired to know analytically the system characteristics related with such parameters.
From these points of view, this paper is intended to synthesize a learning control policy for a linear discrete system (corrupted by additive noises), containing unknown functions of measurable physical parameters.
In this paper, it is assumed that the states of the system are observed in the subintervals of each control stage, to obtain the necessary informations for the on-line function learning.
In this case, due to the additive system noises, the ordinary stochastic approximation method for function learning is not available.
Therefore, the modified stochastic approximation method, which insures the convergence of the on-line unkown function learning in the entire state space, is considered and used to synthesize the suboptimal learning control policy, with the direct application of dynamic programming method.
An another function learning method, which reuses the stored sampled informations by learning the sampling probability characteristics, is presented.
As the examples, some simulation results on digital computer are shown for a two-dimensional control system.View full abstractDownload PDF (1080K) -
Takuya KATAYAMA1969 Volume 5 Issue 4 Pages 386-394
Published: October 30, 1969
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn many cases, natural irregularity waveform W(ξ) of the surfase of a strip is considered to be a sample function from some stationary stochastic process. Suppose the strip is moving at speed υ, signal X(t) obtained by observing it at ξ=0 is given by X(t)=W(υt), so, X(t) is a stochastic process whose simultaneous probability distribution contains υ as one of it's parameters.
Therefore, when observation data X={X(t);t∈T}is given, we_can estimate υ statistically from X and non-contact speed measurement is made possible through it.
In this paper, the author studied on this problem in case where W(ξ) is stationary Gaussian process, whose autocorrelation function is σ2e-α|ξ|, and the observation is time-discrete and subject to diserete white noise, that is,
X={xi=W(iυΔ)+ni;i=1, …, N}, E[ni;nj]=ρ2δij
Maximum likelihood estimate is adopted as υ, estimate of υ, and Cramer-Rao bound is used to evaluate estimation error assuming sample size N is very large.
From this analysis it is made clear that υ is a fairly good estimate in the ordinary measurement conditions.
As for the effect of observation noise {ni}, increase of estimation error ratio (to noiseless case) due to the noise is almost equal to noise-to-signal ratio, therefore, noise is not so harmfull if it is small.
To the contrary, if parameters of the probability distribution of irregularity and noise are unkown, estimation error becomes many times larger than in case where they are known.
It is also painted out that determination of sample interval Δ is very important in the actualsituation. That is, if observation time NΔ is constant, decrease of Δ makes the estimation error small but at the same time it also makes our mathematical model of the irregularity not appropriate for the actual one whose waveform is differentiable.
Although our analysis is made on some special model and therefore not necessarily applicable to actual cases, some insight into the statistical method of non-contact speed measurement is obtained.View full abstractDownload PDF (1104K) -
Hitoshi SASAI1969 Volume 5 Issue 4 Pages 395-399
Published: October 30, 1969
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn computing a numerical solution of a partial differential equation, an approximating method in which differential operator is replaced by a difference operator is widely used.
A similar approximating procedure is commonly used in calculating an optimal control in systems with distributed parameters.
Convergency conditions of a sequence of approximating solutions for the final value and tracking problems of distributed parameter systems are given in this paper.View full abstractDownload PDF (590K)
- |<
- <
- 1
- >
- >|