巻号一覧
35 巻, 6 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
- >
- >|
-
福島 省吾, 新井 豪, 森川 大輔, 下田 宏, 吉川 榮和1999 年 35 巻 6 号 p. 699-707
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーA new head-mounted-display, named as Eye-Sensing Head Mounted Display (ES-HMD), has been developed which can not only provide the user with virtual environment through visual and auditory stimuli conventionally but also monitor the user's both eyes and calculate ocular information such as pupil size, eyeblink or eye gaze point. The obtained ocular measures will be utilized for analysis of its relation to peculiar ocular characteristics under various binocular visual conditions as well as for eye gaze interaction to the display.
The specifications of the developed ES-HMD are described first, followed by the discussion on the characteristics of pupil indices measured by ES-HMD; it is shown by measuring pupil that pupil sphere has better resolution to indicate pupil size than pupil diameter. A new calibration method for the eye gaze point on the display is also presented, which has better performance than conventional linear interpolation method.抄録全体を表示PDF形式でダウンロード (4878K) -
佐藤 浩志, 奥田 日出治, 渡辺 嘉二郎1999 年 35 巻 6 号 p. 708-714
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーIt has been experimentally known that under the certain conditions, the police whistle and its components generate the sound whose frequency is linearly proportional to the amount of blowing. Among the whistle itself and its components, the cylindrical cavity, i.e, one of the whistle components, generates the sinusoidal sound whose frequency is proportional to the flow speed in the wide flow range. I.e., the cylindrical cavity of a police whistle has the function of an excellent flowmeter of frequency detection type.
But, it is not clarified how fluid behaves inside the cylinder of the whistle when whistling, nor why the sound generates. This paper is aimed at describing to clarify these phenomena occurring in the cylinder by the fluid flow visualization experiment and the simulation.
At first, experimental visualization results that show only the part of fluid flow phenomena were compared with the corresponding simulation results. Both results show the same characteristics.
Next, the details that cannot be examined by experiments but can be obtained by the simulations were analyzed.
Finally it was clarified that the cause of the sound generation is due to the cyclic returning of fluid to the inside from the cylinder outside at the outlet part of the cylinder.抄録全体を表示PDF形式でダウンロード (3612K) -
村岡 茂信1999 年 35 巻 6 号 p. 715-722
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーMany force sensors used in robots use strain gauges. Because they have an analogue output, these force sensors are readily affected by electrical noise. They also often require amplifiers because of the low signal level, low pass filters to decrease electrical noise and A-D converters to interface with computers. On the other hand, force sensors which use quartz resonators are robust to electrical noise since the output is a frequency shift in response to a change in an external force. The frequency output of the force sensor can be easily converted into a digital signal through a frequency counter. This type of force sensor also has quick response, high sensitivity, high resolution and wide bandwidth. Therefore, it could be used in an environment with a high electrical noise background and where a quick response is needed. Although force sensors with quartz resonators have many advantages, as mentioned above, such sensors have not yet been applied in measuring dynamic force such as in a grasping force sensor. In this paper, first, mechanical strength of quartz resonators mounted in a robot finger was examined experimentally to static and impulsive forces. Secondly, the relationship between the external force and frequency shift of quartz resonators is discussed. Thirdly, a differential method applied to a robot sensor using quartz resonators is proposed. Finally, the response of the force sensor to an oscillatory force is discussed. On the basis of the aforementioned, a sensor for three components of grasping force and position using three pairs of quartz resonators is proposed. The force sensitivity of the sensor is independent of the grasping position on the finger. Performances of such a force sensor was examined experimentally.抄録全体を表示PDF形式でダウンロード (4575K) -
宮崎 孝, 細江 繁幸1999 年 35 巻 6 号 p. 723-731
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーIn this paper we consider a J-spectral factorization associated with a model matching problem for descriptor systems that contains non-proper systems as a special case. We do not assume that the plant has no invariant zeros at infinity. The existence of solutions to the model matching problem is shown to be equivalent to the J-spectral factorization. The J-spectral factor is constructed by a stabilizing solution of a generalized algebraic Riccati equation that is non-regular type. We also provide solvability conditions of the generalized Riccati equations and a computation method by considering a generalized eigenvalue problem. The results are natural extensions for a standard state space case.抄録全体を表示PDF形式でダウンロード (1218K) -
マイクロトンネル機械への応用相原 伸一, 壁内 輝夫1999 年 35 巻 6 号 p. 732-740
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーWe study a stability property for stochastic hyperbolic equation with free boundary. This model is motivated by a micro-tunneling machine which is used for the conduit installation for communication cables. By using this tunneling machine, a flexible pilot pipe is thrusted into the ground and hence the total length of the inserted pipe becomes a function of time. Furthermore, the thrust force generated by a hydraulic jack is stochastically perturbed. The transverse displacement of the inserted pipe is firstly shown to be governed by a hyperbolic equation with free boundary by using the Hamilton principle. Secondly, the existence of a solution of this equation is proved. For checking the buckling phenomena the stochastic stability is studied. Two digital simulation examples are demonstrated.抄録全体を表示PDF形式でダウンロード (1461K) -
長棟 亮三, 山本 茂1999 年 35 巻 6 号 p. 741-747
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーTo construct a model set is inevitable for designing a controller based on robust control theory. In this paper, we will propose one method of model set identification. The model set treated in this paper is a discrete-time state space model set with time-varying parametric uncertainties measured by some kinds of norms. The proposed method is a slight modification of the subspace method, which is a powerful tool for finding a nominal model in a state space description from I/O data. The minimal model set is obtained via a convex optimization such that it can produce the experimental I/O data from a real system to be identified.抄録全体を表示PDF形式でダウンロード (1400K) -
若佐 裕治, 佐々木 正弘, 谷野 哲三, 片山 徹1999 年 35 巻 6 号 p. 748-753
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーMany control analysis/synthesis problems have been shown to be reducible to bilinear matrix inequality (BMI) problems. In this paper, we present a global optimization algorithm for the BMI problem based on the primalrelaxed dual method. We also modify the algorithm from the viewpoint of computational efficiency. A numerical example is given to illustrate the geometrical interpretation and effectiveness of the proposed method.抄録全体を表示PDF形式でダウンロード (1515K) -
佐伯 正美, 和田 信敬, 土屋 正樹1999 年 35 巻 6 号 p. 754-761
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーA design method of a static anti-windup controller constructed from a left coprime factorization of the original linear controller is proposed. A performance index is derived based on the observation that the overshoot of the step response becomes smaller as the magnitude of the input to the saturation function is made smaller during saturation. The minimization of the performance index is represented as H2 or H∞ control problems with state feedback. The result is extended to a class of controllers which have feedthrough term by applying singular perturbation method. Two parameters contained in the performance index are chosen to improve not only anti-windup property but also robust stability.抄録全体を表示PDF形式でダウンロード (1680K) -
清水 悦郎, 三平 満司, 古賀 雅伸1999 年 35 巻 6 号 p. 762-771
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーThis paper gives a design method of a nonlinear H∞ output feedback controller for linear systems. In order to derive a nonlinear controller, nonlinear weighting functions are used. These nonlinear weights are defined as functions of the observer variable. Under some constraints on nonlinear weights, it is shown that a nonlinear output feedback controller is obtained by solving the algebraic Riccati inequality. This paper also shows one of nonlinear weights which satisfy the constraints.抄録全体を表示PDF形式でダウンロード (1609K) -
石川 将人, 三平 満司1999 年 35 巻 6 号 p. 772-781
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーThis paper investigates necessary conditions for feedback stabilizability of nonlinear control systems, which was originally proposed by Coron (1992). The purpose is to provide some detailed and practical descriptions of the conditions, through assuming that equilibria sets of systems form a regular submanifold of their state space. At first, as a special version of the stabilizability condition, it will be shown that a system is not feedback stabilizable if the number of inputs differs from the dimension of the equilibria submanifold. Secondary, we will show that the stabilizability condition can be reduced in a sense if the equilibria set satisfies a certain intersection condition. As a major application of this result, we will show that the stabilizability condition for affine control systems is independent of the input vectorfields.抄録全体を表示PDF形式でダウンロード (2064K) -
鈴木 良昭, 板宮 敬悦, 鈴木 隆1999 年 35 巻 6 号 p. 782-790
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーThis paper presents a robust scheme for the indirect type of model reference adaptive control (MRAC) system, in which the controller parameter is determined by design calculation from the estimate of plant parameter, in the presence of unmodeled dynamics and measurement noise. This scheme is characterized by the use of adaptive law with dead zone and integral type of fixed compensator. Stability of the MRAC system is analyzed by applying the concept of L2δ norm and the Bellman-Gronwall lemma, based on the properties guaranteed by the adaptive law, and it is shown that the system can behave stably if the modeling error is sufficiently small. Effectiveness of the scheme is demonstrated by the simulation carried out for a plant with second order nominal part.抄録全体を表示PDF形式でダウンロード (1761K) -
森本 滋郎, 山本 由和, 小林 郁典, 古本 奈奈代, 田渕 敏明1999 年 35 巻 6 号 p. 791-799
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーComparisons of the adaptation performance of adaptive algorithms have been studied before. While comparing, conditions of the adaptive algorithms have to be made optimal to be accurate it. This optimality strongly depends on the value of the design variable of the adaptive algorithm, and so it is very important to find out the design variable's optimal value.
Two methods of comparison are possible for the selection of the design variable. One method is based on the analytical approach in which the variation of the unknown time-varying parameters is assumed to be a random walk model and is very slow. In this method, the optimality of the adaptive algorithm is guaranteed. The other way is based on the numerical method in which the assumption of the variation of the parameters is not introduced, though the optimality is not guaranteed.
Considering the restrictions of the above two comparing methods, it is difficult to ensure the optimality of the adaptive algorithms in cases where variation of the parameters is random walk and is not slow or can not be described as a random walk model.
We propose a comparison method among NLMS (normalized least mean square), RLS (recursive least squares) and KF (Kalman filter) based adaptive algorithms based on literatures 5) and 6). This makes the condition of the adaptive algorithms optimal without the distinction of the variation of the parameters. In the numerical experiment, we will show the results of the following three cases that: a) the variation of the parameters can not be described as random walk, b) the variation is not slow and is random walk and c) very slow random walk.抄録全体を表示PDF形式でダウンロード (1860K) -
奥 宏史, 木村 英紀1999 年 35 巻 6 号 p. 800-805
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーThis paper proposes recursive algorithms of state-space subspace system identification for multiple-input, multiple-output (MIMO), finite dimensional, linear time-invariant (LTI) plants. First, a recusive algorithm is derived for noise-free cases. Then, the algorithm is extended to the case with process and measurement noise, which is equivalent to the recursive version of the so-called PO-MOESP method.抄録全体を表示PDF形式でダウンロード (1385K) -
Reiko TANAKA, Kazuo MUROTA1999 年 35 巻 6 号 p. 806-813
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーThis paper deals with the fault-tolerance of systems that are symmetric with respect to the dihedral group Dm. The group Dm generally represents the geometric symmetry of a regular m-gon. We reveal the underlying mathematical mechanism of the loss of controllability for Dm-symmetric systems induced by failures and derive a necessary and sufficient condition to retain the controllability despite a failure. Moreover, we show the minimum number of functioning modules needed to retain the controllability despite the failures. It serves for a quantitative evaulation of fault-tolerance.抄録全体を表示PDF形式でダウンロード (1734K) -
田中 敏幸, 佐久間 聡, 溝口 哲1999 年 35 巻 6 号 p. 814-819
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーThis paper deals with path planning of robot manipulater using spline method and quaternion. Calculations of trajectory and orientation are done independently. In calculating the trajectory, some regions are appointed instead of some points which the endeffector goes through. The trajectory under consideration includes a part of straight line. A curve of the trajectory is assumed as linear term and Fourier series. On the boundary of the curve and straight line, the first derivative of function characterizing the curve has the same value of slope of the straight line. Fourier coefficients are obtained by linear programming so that value of cost function is minimal under those conditions. The pose of the endeffector on the trajectory is calculated by using quaternion. The poses of some points are given by Euler angles; roll, pitch, yaw. Euler angles are converted to quaternions. The poses of other points on the trajectory are obtained by interpolating the given quaternions, using Bezier interpolation method. An example is shown by using skeleton model of robot manipulater.抄録全体を表示PDF形式でダウンロード (1229K) -
櫻井 弘久1999 年 35 巻 6 号 p. 820-822
発行日: 1999/06/30
公開日: 2009/03/27
ジャーナル フリーTo reduce the uncertainty of thermal equilibrium between samples and thermometers, the thermometric fixed points of the superconductive phase transitions of lead and indium are measured by the static method, which have been traditionally measured by scanning temperatures near the phase transitions. The phase transition widths are found to be within 1mK even in the case of the static method. The static method may be suitable to calibrate thermometers at the superconductive phase transitions and to estimeate the uncertainties.抄録全体を表示PDF形式でダウンロード (2468K)
- |<
- <
- 1
- >
- >|