巻号一覧
18 巻, 2 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
- >
- >|
-
梶原 宏之, 古田 勝久1982 年 18 巻 2 号 p. 103-108
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーThis paper presents a numerically efficient algorithm to obtain Kalman's canonical decomposition for a linear time-invariant system. It is shown that eigen-vectors of a matrix QQ'PP', where P and Q' are the controllability matrix and observability matrix respectively, can be utilized to construct the state transformation matrix giving Kalman's canonical decomposition. The minimal realization of the system is easily obtained by the algorithm which means that the basis of the controllable and observable subspace are determined numerically.抄録全体を表示PDF形式でダウンロード (738K) -
大西 謙一, 岩本 俊輔1982 年 18 巻 2 号 p. 109-116
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーThis paper reports a method to detect the location where the characteristic impedance changes partially in the long co-axial cable.
By assuming the cable as the series of uniform sections, the reflected wave for the pulsate voltage applied at the end of the cable is represented with ARMA model with the coefficient of the voltage reflection factor between two adjacent sections.
In the case of low reflection factor, it is possible to detect the location of impedance unbalance point by estimating the voltage reflection factor with Kalman filter application.
In our experiment, we could detect the location of impedance unbalance point within ±5m accuracy at the distance of 2Km from the end of the cable.抄録全体を表示PDF形式でダウンロード (1107K) -
画像の一生成モデル表現出口 光一郎, 森下 巖1982 年 18 巻 2 号 p. 117-123
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーA new type of two-dimensional random field model, named the two-dimensional auto-regressive model, is introduced for stochastic representation of digital images. This model is an extension of the one-dimensional auto-regressive model for time serieses. In the model the gray level of each pixel in the image is represented as a linear weighted summation of gray levels of its neighbour pixels on all sides, added with a white noise.
Statistical properties and spectral representation of the model are discussed. Then the iterative algorithm is proposed to generate the image represented by the model with given parameters. The model identification from a given image by the maximum likelihood method is also discussed. It is shown that the least square error model fitting does not give correct parameters for the noncausal type auto-regressive models. The summation of square residuals multiplied by a correction term is to be minimized to identify the model. Finally, several random images have been generated by the proposed generation algorithm, and from these images the correct parameters have been obtained by the pronosed identification method.抄録全体を表示PDF形式でダウンロード (1513K) -
藤井 省三, 水野 直樹1982 年 18 巻 2 号 p. 124-130
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーThis paper presents two methods for designing a discrete model reference adaptive control system. The controlled plant is a multi-input, multi-output, continuous-time, linear plant having dead time. The Control is performed by a digital computer.
The first method requires no parameters of the plant and the reference model. For the multivariable system having dead time, a system of the nonminimal realzation corresponding to Lüders-Narendra's canonical form is derived. The adaptive control laws and conditions which ensure the stability for the derived system can be obtained in similar procedure as for a single-input, single-output, linear plant.
The second method is in the case where the parameters of the reference model are known. The method makes the design of the reference model easy by assigning a lower order and a diagonal structure compared with the plant, and also makes the adaptive control algorithm simple by using available information of the reference model.
Finally, the results of computer simulation of the adaptive control applied to two-input-output, linear plants are included to illustrate the effectiveness of the proposed methods.抄録全体を表示PDF形式でダウンロード (839K) -
中村 猛見, 鈴木 隆1982 年 18 巻 2 号 p. 131-138
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーIn the model reference adaptive control system (MRACS) by indirect approach, the control input to plant is adaptively synthesized by using the estimates obtained from plant parameter identification. In the conventional schemes, the identification model is constructed corresponding to a plant-output representation, which requires a division by a parameter estimate for synthesizing the control input. This means that the control input may become unbounded. This defect can be overcome by constructing the identification model corresponding to a plant-input representation. This paper presents a design method for discrete type of MRACS based on the above idea.
In this paper, a nonminimal representation with respect to the plant input is first derived and then two schemes-constant gain type and variable gain type-are presented for identifying the parameters of the above representation. A method is then given for synthesizing the control input using the identified values, and the asymptotic staility of the overall adaptive system is discussed by Liapunov's second method. Finally, the computer simulation results are presented and it is demonstrated that the variable gain type of adjustment law yields a superior adaptation response to the constant gain type.抄録全体を表示PDF形式でダウンロード (933K) -
川田 誠一, 嘉納 秀明, 増淵 正美1982 年 18 巻 2 号 p. 139-145
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーAn approximate pole assignment procedure is proposed for some class of first order partial differential equation system whose transfer function has no bounded poles but bounded poles appear when it is feedback controlled.
As this system is equivalent to a system with delay in control, the finite spectrum assignment procedure by A. Manitius et al. seems to be applicable. However it can not be used for this system, since the equivalent system is found not to be absolute controllable.
The original distributed parameter system is approximated by applying the method of weighted residuals. When a feedback control law derived from the approximate system is applied to the original distributed parameter system, the resulting poles are apt to run into unstable region though the locations of assigned poles are in sufficiently stable region. It is found that this instability can be avoided when the two conjugate poles of the approximate system having the largest imaginary parts are fixed in the original locations and the residual approximate poles are appropriately assigned. Moreover, good pole assignments is obtained when the residual poles are assigned in the right hand side of the pole locations of the approximate system.抄録全体を表示PDF形式でダウンロード (901K) -
高田 等1982 年 18 巻 2 号 p. 146-151
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーBy a formal linearization method, a nonlinear dynamic system with nonlinear output is transformed into a formal linear system with linear output. In this paper we study an isomorphism of the two systems. The isomorphism is defined so that the two systems are related by a diffeomorphism between manifolds with the same outputs. An analytic nonlinear system is isomorphic to the formal linear system on the manifold contained in Hilbert space.抄録全体を表示PDF形式でダウンロード (792K) -
長谷川 泰道, 松尾 強, 平野 忠男1982 年 18 巻 2 号 p. 152-159
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーA partial realization problem of discrete time linear representation systems (abb. L.R.S.) is discussed. In reference 7), we have already established the realization theory of an arbitrary black-box with causality condition (equivalently, an input response map a∈F(U*, Y)) by L.R.S.. It contains the uniqueness theorem and existence theorem. We have also investigated finite-dimensional L.R.S. in reference 9). One of the main theorems is realization condition of finite-dimensional L.R.S..
Based upon the preceding results, we investigate in this paper a partial realization problem for finite input-output data of length N of a non-linear black-box (or equivalenty, a partial input response map a∈F(U*(≤N), Y) of length N). The partial realization problem is stated as follows:
“For an arbitrary partial input response map a∈F(U*(≤N), Y), seek a minimum L.R.S. which partially realizes a and want to prove that minimum partial realizations are unique up to isomorphisms.” Existence of minimum partial realizations are trivially presented. It is shown that minimum partial realizations are not unique even up to isomorphisms. To solve the uniqueness problem, we introduce the notion of natural partial realizations. The main contents are following:
1). A necessary and sufficient condition for the existence of the natural partial realizations of a∈F(U*(≤N), Y) is given by the rank condition of finite Hankel matrix of a.
2). The existence condition of natural partial realizations of a∈F(U*(≤N), Y) is equivalent to the uniqueness condition of minimum partial realizations of a modulo isomorphisms.
3). An algorithm to obtain a natural partial realization from a partial input response map a∈F(U*(≤N), Y).抄録全体を表示PDF形式でダウンロード (1132K) -
砂原 善文, 森田 洋二, 安田 寿彦1982 年 18 巻 2 号 p. 160-166
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーThe purpose of this paper is to give a stochastic system which is, in the sense of invariant measure, equivalent to a class of nonlinear deterministic systems exhibiting chaotic behaviors.
First, the f-invariant measure is defined. Then, we shall propose a stochastic system driven by a white Gaussian noise. Taking into account the correction term by Wong & Zakai, the mathematical model is given by the stochastic differential equation of Ito-type. In order to obtain the conditions for guaranteeing the measure equivalence between the deterministic and the stochastic systems, the stationary probability density function of the stochastic system derived above is computed.
Finally, illustrative examples are demonstrated together with digital computer studies representing numerically the relation between the f-invariant measure and the stationary probability density function.抄録全体を表示PDF形式でダウンロード (818K) -
鈴木 茂行, 川人 卓, 沖津 泰1982 年 18 巻 2 号 p. 167-172
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーThe paper proposes a method for measuring torque coefficient (CT) of a wind turbine as a function of tip speed ratio (λ). In the method, time series data representing wind and turbine speed of a wind power system are used for the measurement. The system may be operating either under constant or time varying wind of nature.. Time integrated values of the wind turbine torque of each time interval are supposed to be calculated from the turbine speed and given system parameters. Equations are derived to calculate corresponding values of λ and CT at each time portion of the input data. Experiments ware carried out to test the validity of the method both under constant and natural wind. The test result show the method will be quite applicable to practical use under constant wind speed. The results also show that, though it is promising, more improvement is necessary in measuring wind speed in the case of natural wind. Especially, a method should be established which would enable the estimation of wind speed at a turbine blade. It would utilize time series wind speed and direction data collected at some points near the turbine blade.抄録全体を表示PDF形式でダウンロード (759K) -
富永 雅樹, 木下 武雄1982 年 18 巻 2 号 p. 173-179
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーThe soil-water is a major factor not only in the natural disaster such as landslides, but also in the preservation of the environment such as ground water recharge. Therefore, the soil-water movement is one of the urgent topics to be investigated in these fields. In the present paper, the methods of measuring soil-water velocity insite using the sinusoidal heat signal are developed.
First, the energy equation of flowing fluid is solved under the condition that the point heat source is heated sinusoidally. From the solution, two methods are developed, the amplitude-method and the phase-difference-method. The direction and the absolute value of the velocity are obtained by the former method. By the latter method, though only the absolute value of the velocity is obtained, deviation of the thermal diffusivity can be eliminated.
Second, the energy equation of the porous media as a model of the soil is developed introducing the two media, the flowing medium and the stationary medium, which play important roles in the conduction of heat. The obtained energy equation is similar to that of flowing fluid, so the fluid velocity in the porous media can be measured by the similar way.
Finally, the experiment of measuring the water velocity which is about 0.1-2.0mm/s in porous media is conducted by the phase-difference-method. The results agreed with the theoretical consideration, and the availability of the method to the soil-water velocity in site is confirmed.抄録全体を表示PDF形式でダウンロード (1541K) -
大山 尚武, 重田 清子1982 年 18 巻 2 号 p. 180-185
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーThe detector using fine photoelectric elements was developed to measure the spatial distribution of vehicle velocities on the flow of traffic. To test the equipment and to investigate possible advantages for traffic control, an installation was made on Route 15 in Tokyo. The field test was carried from March 1978 to September 1980. This paper describes the results of this test. Concluding remarks are as follows. (1) The detector has stably and correctly worked for two years without a repair and a regulation. (2) The error in measuring vehicle velocities is less than ±3km/h. (3) As the new area detection capability, the detector is able to estimate the vehicle density in the detection area.抄録全体を表示PDF形式でダウンロード (2200K) -
佐野 昭1982 年 18 巻 2 号 p. 186-192
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーThe statistically optimum receivers give a good performance in binary data transmission systems subject to severe intersymbol interference and statistical noise, but they become too complex to be practical. This paper describes the new practical detector which can be derived by introducing the idea of the decision feedback to the maximum likelihood (ML) bit estimation, and which retains the implementable simplicity like conventional decision feedback equalizers (DFE). The proposed detector has the multilayered structure as if the multiple conventional DFE's are arranged in a parallel form. A method is also developed to estimate the error probability of the detector, and its analysis and simulation results show that, by increasing the level of the multilayers and the tap number of the feedforward filter, the proposed scheme can ahcieve the error rate performance very near to that of the ML bit estimation.抄録全体を表示PDF形式でダウンロード (852K) -
広瀬 茂男, 岩崎 廣之, 梅谷 陽二1982 年 18 巻 2 号 p. 193-200
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーIt is generally recognized that the walking machine will make superior off road vehicle to conventional wheeled or tracked vehicles. The high mobility of walking vehicle come from its intelligent functions to select gait adaptive to the terrain. If the legged vehicle is without the intelligence and the legs are driven in fixed sequence, it can not even maintain stable motion in rough terrain. Thus the intelligent gait control problem is said tc be crucial in the developement of legged vehicle. But only a few studies have been performed on the intelligence of hexapod walking and not a single study have been implemented on quadruped walking. The difficulties to construct the hardware of walking vehicle itself might have been the greatest obstacles for it. The authors have already constructed a quadruped walking vehicle (Perambulating Vehicle Mark II, abbreviated as PV II) and successfully demonstrated its fundamental mobilities. To set the PV II as a object model this paper makes the first attempt to analyse the intelligent gait control problem. The paper includes a complete problem formalization, clarification and analysis of the gait control algorithm, and the evaluation of the proposed algorithm in terms of a computer simulation study.抄録全体を表示PDF形式でダウンロード (1797K) -
誘導電動機速度制御系土谷 武士1982 年 18 巻 2 号 p. 201-203
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーPDF形式でダウンロード (317K) -
安田 一則, 平井 一正1982 年 18 巻 2 号 p. 204-205
発行日: 1982/02/28
公開日: 2009/03/27
ジャーナル フリーPDF形式でダウンロード (276K)
- |<
- <
- 1
- >
- >|