計測自動制御学会論文集
Online ISSN : 1883-8189
Print ISSN : 0453-4654
ISSN-L : 0453-4654
39 巻, 5 号
選択された号の論文の12件中1~12を表示しています
  • 田原 鉄也, 新 誠一
    2003 年 39 巻 5 号 p. 425-431
    発行日: 2003/05/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper concerns properties of wavelet transform of a correlation function from the viewpoint of spectrum analysis. First of all, a basic characteristic is derived as a relation between the wavelet transform and the spectrum density. The relation shows that the wavelet transform can be treated as an estimate of the corresponding spectrum density if an analyzing wavelet is chosen appropriately. Its bias from the true value is also obtained for this case. Moreover, a selection of an analyzing wavelet is considerd to reduce the bias. This consideration provides validity of using the Gabor function as an analyzing wavelet for applications of wavelet transform of a correlation function.
  • 田中 正吾, 山田 実
    2003 年 39 巻 5 号 p. 432-440
    発行日: 2003/05/31
    公開日: 2009/03/27
    ジャーナル フリー
    Nondestructive inspection by the electromagnetic wave (Radar) conventionally presents shaded images on the magnitude variation of the amplitudes of reflected waves. Therefore, identification of cracks and cavities in concrete structures is marginally successful in the viewpoint of the reliability and the accuracy. This is mainly because the conventional method aims a qualitative diagnosis of the anomalies based on inspector's eyes.
    This paper thus proposes an effective inspection procedure which can identify the locations of cracks and cavities quantitatively in concrete structures (and also the thickness of the cavities). It is based on a signal propagation model (i.e., a stack model of reflected waves) and a pattern matching strategy. Results show that the proposed procedure enables an accurate and a reliable identification of cracks and cavities in the structure.
  • 岡田 徳次, 高橋 真之, 佐藤 秀幸
    2003 年 39 巻 5 号 p. 441-448
    発行日: 2003/05/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper proposes a simplified measurement principle of an audio sound source localization in a 3-D space. Such kind of differences like the transferring time, phase shift, and level of the sound wave are treated for the measurement in general. Our approach is to use the differences of sound levels by taking the advantage of the condenser microphone's unidirectional sensitivity characteristics so that the localization is performed by simple signal processing. Basically the sound levels are collected at two sensor towers which are separated with a certain distance. Each tower has four microphones located at tetrapod positions for monitoring the levels from all directions. We show the calculation procedure for extracting the location of the sound source. We made an experimental set and collected sound data for verifying the measurement principle. The results show that the directional and positional errors of the sound source are less than 0.22[rad] and 1.2[m] at the distance 5[m], for instance. Error analysis shows that these data are reasonable. Limitation of measurement accuracy improvement is also described.
  • 児玉 一裕, 袈裟丸 智英, 安達 元之, 神村 一幸, 黒須 茂
    2003 年 39 巻 5 号 p. 449-454
    発行日: 2003/05/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper concerns the development of an entirely new force sensor called Gyroscopic Force Measuring System (simply called GFMS) for measuring a force vectorially. In a previous paper, the dynamical characteristics and the error analysis of the GFMS for measuring a force vector in 3-dimensional space were examined using numerical simulations. The results of this work are directly applicable to design and construction of the GFMS. In this paper, the GFMS is constructed by a gyro-rotor using a miniature rate-gyro for aircraft instrument and mechanical parts together. The feedback gains are selected somewhat arbitrarily, but the fundamental experiments show that the force of less than 0.3N can be measured by the GFMS. The results suggest that the GFMS is satisfactory for measurement of a small force range useful for air-flow distribution in an air-conditioned room as an example of expected applications. No doubt the experimental results are to be released in a future development.
  • Xiaohong JIAO, Tielong SHEN, Katsutoshi TAMURA
    2003 年 39 巻 5 号 p. 455-462
    発行日: 2003/05/31
    公開日: 2009/03/27
    ジャーナル フリー
    In this paper, the robust feedback control problem is investigated for a class of nonlinear systems with uncertain input dynamics and unknown control direction. The uncertainty is restricted to be minimum-phase. First, by constructing a positive definite function guaranteeing the Barbalat's condition, a robust adaptive stabilizing controller is derived for the system unforced by disturbance. Then, a robust adaptive control law is designed for the system forced by disturbance such that the closed-loop system is convergent and the robust L2-gain from disturbance to penalty signal is less than a given level.
  • 藤岡 久也, 伊藤 健作
    2003 年 39 巻 5 号 p. 463-468
    発行日: 2003/05/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper proposes a design method for NCSs (networked control systems), where plant and controller are linked through a serial communication network. The network has limitted capacity and control inputs and measured outputs are updated/sampled partially at each step. Assuming that the controller-plant communication is periodic, the design problem is formulated as one for sampled-data feedback systems with periodic discrete-time components. A necessary and sufficient condition for existence of discrete-time periodic controller is given in terms of LMIs, and a controller construction algorithm is derived. The proposed controller (if exists) is stabilizing and sub-optimizes the L2-induced norm of the resultant NCS.
  • 東 俊一, 井村 順一
    2003 年 39 巻 5 号 p. 469-478
    発行日: 2003/05/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper discusses the controllability problem for a class of continuous-time piecewise affine (PWA) systems with the sampled-data autonomous switching, which we call here the sampled-data PWA systems. First, the four concepts of controllability are introduced based on the aspects of the continuous/discrete state and of the continuous/discrete time, and some inclusion relations among them are discussed. Next, easily checkable, (necessary and) sufficient conditions for the system to be controllable are derived.
  • 加藤 久雄, 不破 勝彦
    2003 年 39 巻 5 号 p. 479-486
    発行日: 2003/05/31
    公開日: 2009/03/27
    ジャーナル フリー
    In many years, large amount of researchers in control engineering community have been attracted by so-called servomechanism problem, such as robust servoproblem, optimal servoproblem. For nonsquare plant such that the number of inputs of plant is than that of outputs, the nonuniqueness of the steady-state responses in servosystems has been discussed.
    In this paper, by using such nonuniqueness, we consider the optimization problem of the steady-state responses of inputs and states. A theoretical framework different from the conventional studies is that the criterion to be optimized is the quadratic form with respect to the amplitude of steady-state inputs and states.
  • 北田 宏
    2003 年 39 巻 5 号 p. 487-493
    発行日: 2003/05/31
    公開日: 2009/03/27
    ジャーナル フリー
    A controller design technique for periodic disturbance rejection is proposed. The form of the controller has a free parameter transfer function similar to the Youla parametrization, and it enables to separate the general property of sensitivity function of closed loop system and detailed design of the periodic disturbance rejection performance. A design method of the parameter function is given using the solution of the model matching problem. Numerical examples of the sensitivity function for the mold level control of continuous caster shows the flexibility of the design. The performance of the disturbance rejection is shown by simulation results of mold level control and an experimental result.
  • 柴田 克成, 上田 雅英, 伊藤 宏司
    2003 年 39 巻 5 号 p. 494-502
    発行日: 2003/05/31
    公開日: 2009/03/27
    ジャーナル フリー
    In this paper, a concept of individuality and sociality is introduced as a method to avoid conflicts of individual interests in multi-agent systems. It is considered that each agent has its individuality when the conflicts are resolved by making its own mapping from the sensory input to the action output. On the other hand, each agent has sociality when the conflicts are avoided by some common input-output mapping, which is commonly called rules. A conflict avoidance task in which passengers are getting on and off a train are taken as an example, and the emergence processes of both behavioral characters are explained. Furthermore, it is shown that the differentiation of the agent into one of them is adaptively realized by reinforcement learning based on local rewards according to the asymmetry of environment, number of agents, identification of the other agents, or physical ability of agents.
  • 吉本 潤一郎, 石井 信, 佐藤 雅昭
    2003 年 39 巻 5 号 p. 503-512
    発行日: 2003/05/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper presents a model selection method for normalized Gaussian network (NGnet). We introduce a hierarchical prior distribution of the model parameters and the NGnet is trained based on the variational Bayes (VB) inference. The free energy calculated in the VB inference is used as a criterion for the model selection. In order to efficiently search for the optimal model structure, we develop a hierarchical model selection method. The performance of our method is evaluated by using function approximation and nonlinear dynamical system identification problems. Our method achieved better performance than existing methods.
  • Shon MIN KYU, Junichi MURATA, Kotaro HIRASAWA
    2003 年 39 巻 5 号 p. 513-519
    発行日: 2003/05/31
    公開日: 2009/03/27
    ジャーナル フリー
    Neural networks with local activation functions, for example RBFNs (Radial Basis Function Networks), have a merit of excellent generalization abilities. When this type of network is used in function approximation, it is very important to determine the proper division of the input space into local regions to each of which a local activation function is assigned. In RBFNs, this is equivalent to determination of the locations and the numbers of its RBFs, which is generally done based on the distribution of input data. But, in function approximation, the output information (the value of the function to be approximated) must be considered in determination of the local regions. A new method is proposed that uses LVQ network to approximate functions based on the output information. It divides the input space into regions with a prototype vector at the center of each region. The ordinary LVQ, however, outputs discrete values only, and therefore can not deal with continuous functions. In this paper, a technique is proposed to solve this problem. Examples are provided to show the effectiveness of the proposed method.
feedback
Top