巻号一覧
42 巻, 1 号
選択された号の論文の15件中1～15を表示しています
- |<
- <
- 1
- >
- >|
-
力学変数の変数変換森田 晋, 大塚 敏之2006 年 42 巻 1 号 p. 1-10
発行日: 2006/01/31
公開日: 2009/03/27
ジャーナル フリーRecently, many studies have been carried out on determining the model of human motor control. Proposed criteria are based on physiological observations. In this paper, a novel motion generation method is proposed purely based on a mechanical principle, Hamilton's principle, in contrast to the former studies. The method is derived through expressing the generalized coordinates as a time differentiation of variables that are its time integration. This preserves the structure of Hamilton's principle, which is expected to account for natural appearance of motion. Numerical examples of human arm reaching movement will be shown for comparison with a physiologically studied motor control model and also commonly used optimization methods.抄録全体を表示PDF形式でダウンロード (2203K) -
御崎 芳仁, 内田 健康, 東 剛人, 藤田 政之2006 年 42 巻 1 号 p. 11-16
発行日: 2006/01/31
公開日: 2009/03/27
ジャーナル フリーWe discuss a state feedback H∞ control problem for a simple networked control system with one linear plant and one controller connected by sensor and actuator channels with time-varying delays under the two types of information structure on the time delays. We approach to this problem by adopting architecture of a predictive control for a controlled output, and characterize the predictive control by using parameter-dependent linear matrix inequalities. We also discuss an output feedback case.抄録全体を表示PDF形式でダウンロード (1381K) -
Osamu KANEKO, Shotaro SOUMA, Takao FUJII2006 年 42 巻 1 号 p. 17-25
発行日: 2006/01/31
公開日: 2009/03/27
ジャーナル フリーIn 19) and 20), the authors proposed the tuning method of a controller by using only one-shot experimental data, which is referred to as Fictitious Reference Iterative Tuning (which is abbreviated to FRIT). FRIT requires only one-shot experimental data for an off-line non-linear optimization in order to obtain the optimal parameter minimizing the error between the desired output and the real one, so this is a useful and effective method in the sense that the cost and time for tuning of a controller can be drastically reduced. In this paper we expand FRIT into the two-degree of freedom control scheme. We show that our FRIT works effectively in the two-degree of freedom control systems and the minimization of the error in the fictitious area by using an off-line non-linear optimization corresponds to that in the real area. Moreover, by focusing the role of a feedforward controller in the two-degree of freedom control scheme, we also provide a new and facile closed loop system identification as one of the applications of the obtained result. Finally, we give an experimental result in order to show the validity of the proposed method in this paper.抄録全体を表示PDF形式でダウンロード (1799K) -
喜種 奈美, 中村 文一, 山下 裕, 西谷 紘一2006 年 42 巻 1 号 p. 26-34
発行日: 2006/01/31
公開日: 2009/03/27
ジャーナル フリーSontag's controller without input constraint is an inverse optimal regulator and holds the robustness for disturbances. When there are input constraints, however, any inverse optimal problems have not been clarified and any controllers in previous works do not guarantee the robustness. In this paper, we propose inverse optimal controllers for a nonlinear system with an input constraint under the assumption that a local control Lyapunov function is given. First, we construct two inverse optimal regulators. Then, we design robust controllers without reduction of a domain of attraction. Moreover, we illustrate the robustness of the proposed controller by computer simulation.抄録全体を表示PDF形式でダウンロード (1878K) -
佐伯 正美2006 年 42 巻 1 号 p. 35-39
発行日: 2006/01/31
公開日: 2009/03/27
ジャーナル フリーA set of stabilizing PID gains is examined by Bounding Crossing Theorem. First, the previous studies on the boundary of the stabilizing set are summarized where the conditions are represented by the frequency response of the plant. Next, a new method of determining the change of the number of unstable roots across the boundary of the stability set is given, by which the number of unstable poles can be easily determined for PI and PD planes.抄録全体を表示PDF形式でダウンロード (1264K) -
十河 拓也2006 年 42 巻 1 号 p. 40-46
発行日: 2006/01/31
公開日: 2009/03/27
ジャーナル フリーThis paper investigates the design problem of an infinite-time preview feedforward controller from the viewpoint of the model-matching problem based on the two-sided Laplace transform. It is shown that the so-called preview-based stable inversion for linear time-invariant systems is equivalent to the noncausal solution to the model-matching problem. The preview feedforward controller can be easily discussed in the frequency domain where the design method for the optimal feedback controller has already been established. Moreover, with the same approach as stable inversion, we derive a transfer-function representation for the algorithm of adjoint-based iterative learning control. This indicates that the algorithm can be applied to infinite-dimensional systems such as flexible structures. This is demonstrated through an experiment on the tip control of a flexible arm.抄録全体を表示PDF形式でダウンロード (1676K) -
藤井 信治, 泉田 啓, 真野 修輔2006 年 42 巻 1 号 p. 47-53
発行日: 2006/01/31
公開日: 2009/03/27
ジャーナル フリーThe Q-learning is one of typical reinforcement learning methods. Since the Q-learning requires huge amounts of time to solve a problem, this study proposes acceleration methods. This study introduces two approaches based on iteration methods of the dynamic programming to accelerate the learning. One is to use Robbins-Monro estimation of the state transition probability model. The other is application of iterative solving methods for an inverse matrix, e.g., Jacobi's method, Gauss-Seidel's method, SOR method, etc. Those allow us to determine an appropriate learning factor. Numerical simulations show that the proposed methods are more efficient than the Q-learning.抄録全体を表示PDF形式でダウンロード (1910K) -
凌 元錦, 鈴木 和彦, 幸田 武久2006 年 42 巻 1 号 p. 54-61
発行日: 2006/01/31
公開日: 2009/03/27
ジャーナル フリーIn chemical and nuclear power plants, the safety policies have been established according to the consideration of independent protection layer (IPL). This paper studies risk reduction using the IPL consisting of n layers. For general failure distribution function of plant and IPL, we proposed a general algorithm to estimate the expected numbers of normal trips and accidents caused by IPL failure. The results of this algorithm gave agreement with those had been concluded by Markov analysis in the case of exponential distribution. Moreover, the expected numbers of normal trips and accidents are derived analytically when n≤3. Furthermore, we provided a new quantification method for assessing the risk reduction by the use of IPL. The method is illustrated through application to the risk assessment of nitrobenzene synthesis batch plant.抄録全体を表示PDF形式でダウンロード (1773K) -
渕川 康裕, 黒木 秀一, 松尾 克宏, 宮本 俊一郎, 西田 健2006 年 42 巻 1 号 p. 62-69
発行日: 2006/01/31
公開日: 2009/03/27
ジャーナル フリーA vision-based navigation method for a mobile robot is presented. The present method uses a number of guideposts placed along a navigation path in order for a mobile robot to self-localize and plan a path to go. Precisely, an artificial landmark and a bar-code are put on a face of the guidepost, where the former is used for self localization and the latter for indicating a command for navigation. For self-localization and scaning a bar-code, a visual measurement system with a single camera is used for taking addvantages of simpler and less cost method than stereo-vision system. Although the visual measurement method employed requires a couple of seconds for processing an image, we present a method of path planning for moving smoothly. We have verified that the navigation method works well by means of experiments using actual robot.抄録全体を表示PDF形式でダウンロード (2530K) -
伊藤 友孝, 弓達 孝治, 松井 隆2006 年 42 巻 1 号 p. 70-79
発行日: 2006/01/31
公開日: 2009/03/27
ジャーナル フリーThe consideration on the human characteristics is a key issue to design a practical human-machine cooperative system and to improve the system maneuverability and acceptability because the human operator is one of the most important components that govern the performance of the control system. In this research, we consider a design problem of a network teleoperation system especially and, in this paper, the human characteristics in a feedback control loop and the effect of the system characteristics and communication time delay are experimentally investigated. An application to the stabilization assistance system for the network tele-operation is designed in consideration of the human operation characteristics.抄録全体を表示PDF形式でダウンロード (4376K) -
小川 浩司, 中坊 嘉宏, 向井 利春, 安積 欣志, 大西 昇2006 年 42 巻 1 号 p. 80-89
発行日: 2006/01/31
公開日: 2009/03/27
ジャーナル フリーOur goal is to make an artificial muscle, ionic polymer metal composites, swim with snake-like bending motion. It is easy to miniaturize the artificial muscle because of its simple structure. In addition, the snake-like motion sweeps a smaller area than simple bending motion. Thus it is suitable to apply to future coming swimming robots in thin tubes or in blood vessels, etc. To realize the snake-like bending motion, we cut the surface of the artificial muscle by laser and made insulated patterns on it. By applying voltage with different phase, we can control each segment individually. We have developed a variety of motions including snake-like motion from this patterned artificial muscle. We have also measured the impelling force generated by its snake-like motion. By changing phases and frequencies of voltages which are input to each segment, we can control the speed and the direction of the locomotion. We succeeded in making the muscle swim forward and backward by finding the optimal phase and frequency.抄録全体を表示PDF形式でダウンロード (6941K) -
古田 一雄, 森野 耕平2006 年 42 巻 1 号 p. 90-97
発行日: 2006/01/31
公開日: 2009/03/27
ジャーナル フリーIncreasing demands for participatory approaches of social decision-making is drawing people's interest on public opinion development, but mathematical models used for studying the process so far were insufficient. Dynamics and mechanism of public opinion development therefore were analyzed by multi-agent simulation, which is now a promising approach for analyzing complex. social systems. A simulation model was proposed based on a_decision model that includes both effects of conformity to others and private self-consciousness. Factory that may greatly affect the process were also considered: mass media, personal variation, and bias in true or false of assertion under argument. The result of simulation provided useful insights on the process of public opinion development, e.g., the society splits-into communities of positive and negative opinions, community size increases with increasing conformity to others, and public opinion becomes one-sided by the influence of mass media or bias in true or false of assertion. These findings are both good news and warning for promotion of consensus development in society.抄録全体を表示PDF形式でダウンロード (3005K) -
山中 一雄, 菊池 浩司, 齊藤 充行2006 年 42 巻 1 号 p. 98-100
発行日: 2006/01/31
公開日: 2009/03/27
ジャーナル フリーA solution is presented of a state estimation problem in linear systems driven by white noise. No statistical information is assumed on the observation noise except finiteness of mean power, while the driving noise is characterized by (known) constant power spectral density. The problem is considered as a game between the estimator and the noise by means of introducing a cost function that evaluates the mean-square estimation error subtracted by the mean power of the observation noise. Based on the positive definite solution of a certain Lyapunov equation, a state estimator of Kalman-Luenberger form is obtained as minimax solution of the game.抄録全体を表示PDF形式でダウンロード (435K) -
2006 年 42 巻 1 号 p. 101a
発行日: 2006年
公開日: 2009/03/27
ジャーナル フリーPDF形式でダウンロード (63K) -
2006 年 42 巻 1 号 p. 101b
発行日: 2006年
公開日: 2009/03/27
ジャーナル フリーPDF形式でダウンロード (63K)
- |<
- <
- 1
- >
- >|