巻号一覧
33 巻, 11 号
選択された号の論文の12件中1～12を表示しています
- |<
- <
- 1
- >
- >|
-
上野 直広, 金子 真1997 年 33 巻 11 号 p. 1053-1058
発行日: 1997/11/30
公開日: 2009/03/27
ジャーナル フリーThis paper proposes the Self-excited Dynamic Active Antenna (SDAA) that can detect a contact location between an insensitive flexible beam and an object through observation of the fundamental natural frequency of the beam in contact with the object. The main part of SDAA is composed of a flexible beam, one permanent magnet and two electromagnets. The permanent magnet is attached to a part of the beam and two electro magnets are positioned at both side of the permanent magnet with a small gap. These magnets contribute to not only exciting the motion but also changing the boundary condition of the beam. We analyze the rerationship between the displacement of the permanent magnet and the induced force, and show the linear spring-like behavior of the magnetic force. We also show that the contact position is uniquely determined by utilizing two fundamental natural frequencies before and after changing the boundary condition, whereas both the fundamental and the second natural frequencies are necessary for localizing the contact point under the conventional type of Dynamic Active Antenna. We presents the basic working principle of SDAA and show some experimental result to verify the idea.抄録全体を表示PDF形式でダウンロード (2198K) -
川崎 晴久, 清水 年美1997 年 33 巻 11 号 p. 1059-1065
発行日: 1997/11/30
公開日: 2009/03/27
ジャーナル フリーThis paper presents a symbolic analysis of the base parameters for robots performing kinematic constrained task and/or having closed-chain mechanism by means of the completion procedure in polynomial ideal theory. The regresser of the robot dynamics is represented as a matrix of multivariate polynomials and reduced to the “normal form” besed on Buchbergers' algorithm by constructing reduced Grobner bases from kinematic constrained equations. The linear independence of column vectors of the reduced regresser is examined by the Gauss-Jordan elimination method. The original dynamic parameters are regrouped and some of them are eliminated based on the results. The feature of this method is that kinematic constrained equations are not needed to be solved explicitly. This method derives all the base parameters systematically in theory. Two examples are presented to illustrate the merits of the new method.抄録全体を表示PDF形式でダウンロード (1269K) -
平田 健太郎, 山本 裕, Allen TANNENBAUM, 片山 徹1997 年 33 巻 11 号 p. 1066-1071
発行日: 1997/11/30
公開日: 2009/03/27
ジャーナル フリーThis paper derives a rank type formula for the computation of the singular value equation arising in the H∞ problem with infinite-dimensional unstable plants. It is based on an extension of the skew Toeplitz approach for the stable plant case. Similarly to the stable plant case, parametric representations of the singular value equation plays an essential role. An example is given to illustrate the result.抄録全体を表示PDF形式でダウンロード (1438K) -
大久保 重範1997 年 33 巻 11 号 p. 1072-1080
発行日: 1997/11/30
公開日: 2009/03/27
ジャーナル フリーIn the past time several designs of nonlinear regulator were proposed. Infinite expansion of state series was used in these methods, so convergence of the infinite series was difficult problem. This paper shows the rigorous design of nonlinear regulator using genetic algorithms. When the controlled object has (2N-1) order of state, and peformance function has quadratic form of power state vector of which polynomial order is (2N-1), the optimal control law can be obtained as the feedback of (2N-1) order power state vector. The coefficent matrix of feedback is constracted by a solution of an extended Riccati equation. An extended Riccati equation can not be solved mathematically because the number of variable is less than the number of equation. Genetic algorithms are used in searching method of the solution of an extended Riccati equation. Using the solution of an extended Riccati equatiuon, a Lyapunov function of quadratic form of N order power state vector can be constracted, so a nonlinear regulator which has global stability can be obtained.抄録全体を表示PDF形式でダウンロード (1460K) -
高原 茂幸, 宮本 定明1997 年 33 巻 11 号 p. 1081-1086
発行日: 1997/11/30
公開日: 2009/03/27
ジャーナル フリーA problem of optimizing locations of many rectangular pieces for efficiently utilizing a sheet is considered. The object is to minimize the amount of waste. This problem which is called the nesting problem is important in industrial applications, but it's NP hard. This problem is considered in two stages of the ordering the pieces and the selecting locations of the pieces using the order. In this paper, we propose a new greedy algorithm, called ridgeline method for selecting the optimal locations in the second stage. The order of the pieces strongly affects the performance of the method. Therefore, methods of meta-heuristics, i.e., local search, simulated annealing, tabu search and genetic algorithms, to optimize the order in the first stage are considered. Two options in genetic algorithms are used. These methods of meta heuristics as well as random ordering are compared using simulations experiments. The result shows that the simulated annealing is better than the other meta-heuristics.抄録全体を表示PDF形式でダウンロード (1663K) -
有澤 正樹, 和多田 淳三1997 年 33 巻 11 号 p. 1087-1092
発行日: 1997/11/30
公開日: 2009/03/27
ジャーナル フリーIt is necessary to determine the size of a hierarchical neural network before we make the neural network learn some data. Moreover, the structure of a network has a large influence on the basic abilities to realize mapping between input and output and produce the generalization power. Generally, a network structure is determined by trial and error and on the basis of the experience and knowledge of the designers.
In this paper, we propose a Structural Learning Algorithm for a network which begins and carries out learning from a bigger network, which reduces redundant links according to fuzzy reasoning, and removes units based on their contribution. Stress is placed on the following points: It should not be necessary to determine the detailed structure of the network in advance, the proposed algorithm can reduce redundant links and units in the learning process, it enables us to build as compact a structure as possible, and also enables mapping between input and output. Furthermore, the proposed method can reduce the learning time which has been considered an unavoidable cost in the construction of a neural network, and sufficiently decrease its average of squared errors.抄録全体を表示PDF形式でダウンロード (1692K) -
喜多川 健, 本間 経康, 阿部 健一1997 年 33 巻 11 号 p. 1093-1098
発行日: 1997/11/30
公開日: 2009/03/27
ジャーナル フリーWe propose a learning method for recurrent neural networks with dynamics. The core of this method is to keep a complexity of the network dynamics in the vicinity of the edge of chaos. To investigate the properties of the dynamics effectively and explicitly, novel stochastic parameters defined by combinations of the standard parameters such as individual connection strengths and thresholds are introduced, and then relations between complexities of the dynamics and the stochastic parameters are revealed. The standard parameters are changed by the core part based on the relations and also according to the global error measure. Some examples suggest that the method is practical one for temporal supervised learning tasks and therefore the dynamics of the edge of chaos are effectual for learning of the recurrent networks.抄録全体を表示PDF形式でダウンロード (1603K) -
堀内 英一1997 年 33 巻 11 号 p. 1099-1104
発行日: 1997/11/30
公開日: 2009/03/27
ジャーナル フリーThe introduction of robots to innovative tasks requires discovering skills, or knowledge which leads us to the success of the target tasks. This report proposes a general framework for supporting the skill discovery of robotic tasks by human operators; the presented framework expects human operators to complement where computers fail. Diminishing the role of human operator in the current framework provides a perspective toward complete automatization of the skill discovery process.
We define skill as a continuous function from input space with sensory information and time variables to output space with actuator control information. We approximate skill functions with polynomial expressions and reduce the skill discovery to a combinatorial optimization problem among the coefficients of monomials in the expressions. A genetic algorithm is applied to the optimization because of its all-around applicability and its robustness against local optima. The current framework relies on human experts for the strategies of making efficient search, such as partitioning search space into smaller ones and designing better fitness for GA. The procedure to extract basic mechanism explicitly out of the acquired robot behavior is called skill understanding. Our framework enables this procedure because the skill represented by polynomials is partitioned into monomials and which monomials dominate control information can be analyzed by human operators. In order to verify the validity of the proposed framework, two examples are examined; a visual servo problem of catching a flying ball by a robot baseball player and a problem to find open-loop control of a mechanical system with nonholonomic constraints. The first example illustrates the skill understanding procedure and the second one exhibits a unified manner to solve robotic problems with nonholonomic constraints.抄録全体を表示PDF形式でダウンロード (1668K) -
申 鉄龍, 田村 捷利1997 年 33 巻 11 号 p. 1105-1107
発行日: 1997/11/30
公開日: 2009/03/27
ジャーナル フリーAn uncertain nonlinear system has robust H∞ performance if there exists a positive function satisfying Hamilton-Jacobi inequality with the uncertain perturbated functions. It has been shown that a sufficient condition for the existenc of such positive function is as follows: an Hamilton-Jacobi inequality with a scaling function has positive definite solution. This paper shows the necessity of the condition.抄録全体を表示PDF形式でダウンロード (473K) -
森本 滋郎, 山本 由和, 小林 郁典, 古本 奈奈代, 田渕 敏明1997 年 33 巻 11 号 p. 1108-1110
発行日: 1997/11/30
公開日: 2009/03/27
ジャーナル フリーKalman filter based parameter tracking algorithm for linear regression models is developed. Main subject is to present a selection method of design parameter in the algorithm. It is shown that the algorithm proposed is asymptotically optimal in some sense.抄録全体を表示PDF形式でダウンロード (416K) -
線形行列不等式による凸解析アプローチ佐々木 清吾, 内田 健康1997 年 33 巻 11 号 p. 1111-1113
発行日: 1997/11/30
公開日: 2009/03/27
ジャーナル フリーWe show a synthesis method of nonlinear state feedback control for input-affine polynomial-type nonlinear systems. The method consists of having this problem result in convex programming and solving linear matrix inequalities. To demonstrate efficiency of this method, we also show acomputer simulation for abilinear model of a continuous stirred tank reactor.抄録全体を表示PDF形式でダウンロード (465K) -
李 徳宇, 柴田 浩, 藤中 透1997 年 33 巻 11 号 p. 1114-1116
発行日: 1997/11/30
公開日: 2009/03/27
ジャーナル フリーThis paper proposes a method which makes a discrete-time system ASPR by the insertion of supplementary dynamics (i) in feedback with the plant; (ii) in parallel with the plant; (iii) in cascade with the plant, respectively.抄録全体を表示PDF形式でダウンロード (286K)
- |<
- <
- 1
- >
- >|