計測自動制御学会論文集
Online ISSN : 1883-8189
Print ISSN : 0453-4654
ISSN-L : 0453-4654
31 巻, 7 号
選択された号の論文の17件中1~17を表示しています
  • 寺本 顕武, 大村 晃代, 金尾 博之
    1995 年 31 巻 7 号 p. 815-823
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper describes a newly proposed image reconstruction algorithm and experimental results of a 3-D acoustical holographic imaging system which has an array composed of sparsely distributed transducer elements. The new algorithm is based on the Projection Onto Convex Sets (POCS) method. The POCS method allows the addition of convex sets constrained by a priori information to reduce ambiguity in the image reconstruction process.
    By several experiments, it is proved that the concept of the image reconstruction algorithm has following improvements: 1. the artefacts caused by the grating lobes can be reduced under the condition the interspacing of elements is much larger than the wave length, 2. the storage of the entire inverse matrix of the degradation matrix of the imaging system is not required because the new image reconstruction algorithm is based on the localized projection procedures, 3. the instability to the noise caused by the lack of the knowledge of the point spread function can be reduced.
  • 水中3次元撮像システムの開発
    田村 安孝, 石原 知明, 伊藤 渡, 石井 規夫, 久本 修三, 湯浅 肇
    1995 年 31 巻 7 号 p. 824-833
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    An underwater real-time imaging system based on acoustical holography utilizing an encoded wavefront is constructed.
    The imaging system has a coaxial circular array consisting of ultrasound transmitters and receivers. The transmitters are simultaneously driven by Walsh function modulated signals. Transmitted pulses generate an ultrasound encoded wavefront while the echo waveforms returning from distinct points are approximately uncorrelated. Both the transmitted and received beams are synthesized by a numerical decoding and delay-and-sum process.
    The system can reconstruct images with a single transmission, but to reduce artifacts in one-shot images, a pulse repetition method has been proposed in the previous paper.15) This method utilizes a zero cross-correlation property of repetitively transmitted signals.
    In this paper, the prototype imaging system is described. The characteristics of newly developed transducers are demonstrated. The range of the area where the zero cross-correlation holds for given number of the repetitions is derived theoretically.
    The system has 16ch arbitrary waveform generators and 16ch A/D converters. Transmitters and receivers of wide frequency band-width and wide angle of view were developed. A performance test was carried out in a large towing tank. The system's spatial resolution is 5mm longitudinal and 0.3 degrees in azimuth. Images of targets made of 50φsteel pipes were reconstructed in the range of 6m to 110m, and a pipe of 50×50 square cross-section was reconstructed at 170m distance. B-mode cross-sectional images and C-mode focal-plane images are reconstructed and displayed within one second for the targets at a range of 20m; the images being obtained with a single ultrasound pulse transmission. The test also demonstrated that image quality was improved with 4 repetitive transmissions, along the ability reconstruct 3-D images.
  • 劉 康志, 美多 勉
    1995 年 31 巻 7 号 p. 834-843
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    In recent years, linear control theories, especially the linear optimal control theories, have come to be developed in a general feedback framework which includes the models of disturbances and plant uncertainties that are are usually known as weighting functions. In such framework, the stability common to both regulator and servomechanism problems requires that the weighted closed loop transfer matrix be stable in addition to the internal stability of the actual feedback system. Since some weighting functions may be unstable, this stability requirement is not always solvable even when the plant is internally stabilizable. Such stability requirement is termed as comprehensive stability and analyzed in this paper. The necessary and sufficient condition is derived in a extremely general case and stated in terms of the invariant-zeros of the given generalized plant which contains the dynamics of the plant and weighting functions. Also, a class of controllers achieving the comprehensive stability is parameterized. This result is believed to form a foundation for a unified control theory of regulator and servomechanism.
  • 郭 海蛟, 石原 正, 竹田 宏
    1995 年 31 巻 7 号 p. 844-852
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    Application of the partial LTR (Loop Transfer Recovery) technique to design integral controllers for discrete-time non-minimum phase plants is discussed. We focus our attention to the feedback property at the plant input side assuming that unit step computation delay exists in a controller. First, based on the all-pass/minimum phase decomposition of the plant transfer function matrix, we construct the minimum phase state feedback (MSF) integral controller which feeds back the minimum phase state of a plant. For the output feedback case, we construct the minimum phase estimate feedback (MEF) integral controller by replacing the minimum phase state in the MSF integral controller by its estimate generated by the prediction type Kalman filter. We show that the feedback property achieved by the MEF integral controller approaches that of the MSF integral controller as the intensity of a fictitious disturbance inserted between the minimum phase and the all-pass parts tends to infinity. For the performance index consisting of the quadratic forms of the plant output and input, the feedback property achieved by enforcing the conventional LTR technique coincides with that achieved by the partial LTR technique. This fact clarifies the system theoretic significance of the enforcement of the conventional LTR technique. It should be noted that this fact does not make the partial LTR technique obsolete because it provides more design freedom than the conventional technique. A numerical example making use of the freedom is presented.
  • 森岡 弘, 和田 憲造, Asif ŠABANOVIC, Karel JEZERNIK
    1995 年 31 巻 7 号 p. 853-860
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    In this paper Sliding Mode Control (SMC) with estimated equivalent control is presented. A equivalent control in SMC is a continuous control which is introduced to analyze the sliding mode motion on the manifolds. Using a equivalent control as a control input after reaching manifolds is very useful method from a view point of eliminating chattering and analyzing of sliding mode motion on the manifolds. We presents a design method of on-line estimator which estimates a part of equivalent control containing a system's nonlinear term by use of one of a Neural Network's most powerful ability, that is, function approximation. This control algorithm doesn't require abundant information about controlled plant to calculate a equivalent control. The controller is designed to provide sliding mode motion on the selected manifolds in the state space and to satisfy selected Lyapunov function and stability criteria. This resulting control is continuous, so chattering is eliminated. Under the condition of estimation of on-line estimator converging to real value, the motion of sliding mode is equal to that one of SMC with discontinous control.
  • 安定規範とマニピュレータによる補償運動
    黄 強, 菅野 重樹, 加藤 一郎
    1995 年 31 巻 7 号 p. 861-870
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    Many future applications of robotic systems will require that manipulators perform operations while being carried by moving vehicles. However, different from a manipulator fixed on the floor, such a vehicle-mounted mobile manipulator might be unstable or even tip over. Previous work on stability control hardly considers dynamics and environmental disturbances. The stability of a mobile manipulator has a close relation with the vehicle's motion, the manipulator's posture and motion, and the endpoint's force. And it is necessary for a mobile manipulator to have proper stability when it is used for tasks in various environments with disturbances. This paper first discusses concepts about stability such as the stability degree and the valid stable region based on the ZMP (Zero Moment Point) criterion, which can be used as effective stability criteria in controlling a mobile manipulator. Then, as a control scheme for maintaining or recovering stability, the method of ZMP path planning by a stability potential field is presented. Based on the above-mentioned considerations, a motion planning algorithm is then formulated, which controls the manipulator in order to maintain the stability of the whole system while the vehicle is moving along a given trajectory. Finally, the effectiveness of the method is illustrated by computer simulation.
  • 安藤 慎悟, 中村 裕司, 示村 悦二郎, 内田 健康
    1995 年 31 巻 7 号 p. 871-879
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    In this paper, a new method for discrete-time repetitve control is presented. The control law is obtained using a technique of preview control based on LQ optimal control.
    As a similar method, preview-repetitive control (PRC), which is obtained using a technique of model predictive control, has been proposed. And the stability of the PRC system was studied, but explicit conditions for the stability were not derived because the relation between design parameters and the properties of the PRC system is not clear.
    Using optimal preview control instead of model predictive control, an explicit condition for the stability of the closed-loop system is derived. Furthermore, since the proposed feedback system has the properties of LQ optimal control, starightfoward and systematic design procedures can be given.
    It is shown that the tracking error between the output and the periodic reference command converges to zero as time tends to infinity by increasing preview steps. In addition, the relation between weighting matrices of the performance index and preview steps which guarantees the stability is studied. Furthermore, the asymptotic property of the system for weight selection is investigated. Numerical examples for motor position control illustrate the effectiveness of this method.
  • 小林 一行, 宗像 史生, 渡辺 嘉二郎, 村山 理
    1995 年 31 巻 7 号 p. 880-888
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper describes how to realize a highly accurate positioning system by the hybrid vehicle navigation without using the map information. Here, we present a strategy based on the hybrid vehicle navigation by an optical rate gyro, a vehicle speedometer and a differential GPS (Grobal Positioning System). Especially, we present a positioning system that extrapolates the trajectory continuously when the radio waves from the satellites are not available for a long time interval. The Kalman filter is employed to estimate the accurate position from multiple-measurements. How to build the state variable equation in order to combine the informations from the vehicle sensors and the differential GPS and how to obtain the covariance of measurement and system noises are described.
    The validity of the method was examined by experiments carried out by a real automobile on real circumstances. The highly accurate positioning was obtained without the map information.
  • 澤田 祐一, 大住 晃
    1995 年 31 巻 7 号 p. 889-897
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper presents a method of control for suppressing vibration of a flexible rectangular plate which is subject to the distributed random disturbance. We consider that the control force and torques are acting at the edge of the plate (controlled boundary). To obtain the control force and torques, it is prerequisite not for the boundary conditions but for the system dynamics to include these terms. In spite of this requirement, the control terms disappear in our system dynamics. By considering the damping force and torques acting at the controlled boundary, the behavior of the controlled boundary is described by the first-order differential equation. This leads us to express the control terms in the system equation. Employing the finite-dimensional approximation, the state estimator and the optimal controller are constructed. Furthermore the simulation studies are performed.
  • 藤村 茂, 富田 昭司, 桑原 修二
    1995 年 31 巻 7 号 p. 898-907
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    The paper describes attempts to provide a shell for building intelligent systems in consideration of a tight integration of the knowledge representation, the inference mechanism, the methodology of construction and the development environment. The Shell is called HyperAuk: Shell for building Intelligent Systems. It is a knowledge engineer's assistant tool that provides the graphical development environment to support the construction process, that includes planning, editing, debugging and testing phases.
    HyperAuk enhances the object oriented knowledge representation auk: autonomic knowledge representation, we proposed. The methodology behind HyperAuk provides the concept of the top-down planning, the bottom-up construction and the phaseless development, and defines the sequence of its construction process. According to this concept, HyperAuk constructs the development environment to support the whole development process. This environment is also including the function of the graphical user interface construction. HyperAuk provides a uniform framework by generalization of top-down planning from the methodology view. So the environment for top-down planning is able to be used in any cases.
    HyperAuk is now operational using Smalltalk-80, and its environment is constructed using the user interface management system Talkie, we also developed.
  • 本間 経康, 佐藤 光男, 阿部 健一, 竹田 宏
    1995 年 31 巻 7 号 p. 908-915
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    For a control problem of a complicated nonlinear system with unknown dynamics, system identification method is a useful method, because the method makes clear the properties of the system. But this method with usual linear approximation model is not useful enough for the properties of natural phenomena, for instance atmospheric phenomena, earthquakes, etc., and of biological behaviors, for instance brain action, an organic action, etc. When such a complicated system is analyzed, it is not able to make precisely clear the properties of the system by analyzing properties of individual elements separated from the whole system, because such elements of such a complicated system have polysemy. In other words, in such a system, complicated interaction among the elements exerts an influence upon the system behavior.
    For this problem, in this paper, a new model of complicated nonlinear systems with unknown dynamics is proposed, and an adaptation algorithm using the new model is presented. The model consists of a holon network in which the reciprocal action among the elements (holon) exists, and which is able to realize evolution of its performance adapting to its environment by changing network parameters. In this adaptation algorithm, the network structure is reconstructed by autonomous decentralized adaptation of the parameters keeping balance of two properties of holon, which are properties as the whole to the parts and as the parts to the whole. It is shown for some examples that this algorithm leads to a network structure adapting to the environment and that this model has sufficient generalization ability.
  • 王 丹, 稲垣 敏之
    1995 年 31 巻 7 号 p. 916-922
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper proposes three kinds of redundancy configurations for a safety monitoring system which consists of alarm subsystems with human interface of fault-alerting type or safety-presentation type. For each redundant configuration, we evaluate probabilities of spurious shutdowns (SS) and hazardous accidents (HA). We give conditions for which a redundant configuration can have less SS and HA probabilities than cases in which a single alarm subsystem is utilized. Moreover, we prove that a safety monitoring system consisting of safety-presentation alarm subsystems can be optimal in the sense that it can have the least SS and HA probabilities among possible redundant configurations.
  • 藤村 茂, 三竹 治子, 富田 昭司, 高橋 公一
    1995 年 31 巻 7 号 p. 923-932
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    This paper describes user friendly construction methods for batch plant scheduling systems. Many kinds of flexibility handling many sorts, low quantity and high value products are indispensable for a plant operating scheduling system. Sorts of products are changed little by little in demand. For that, the plant may be reconfigured. In such case, the scheduling system must have an ability to customize easily by plant operators. We have already proposed a framework for an actual plant operating scheduling system. As the enhancement of this system, this paper describes a definition method for the automatic scheduling strategies based on the visual programming technology and a manual modification method for a plan that is scheduled by the automatic scheduling mechanism. This System is implemented on Objectworks_??_Smalltalk. It is applied on many actual plants, so in this paper an example is also described.
  • 玉置 久, 西川 〓一
    1995 年 31 巻 7 号 p. 933-940
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    In recent years, as the diversification of demand and the progress of production technology have brought about a wider variety of products, the necessity of scheduling and inventory control is increasing. This paper deals with a job-shop scheduling problem with consideration of the capacities of in-process buffers, and proposes three kinds of methods of modeling the problem. First, the problem is formulated mathematically into a mixed integer programming (MIP) problem, and referring to the MIP, a way of modeling the problem into a disjunctive graph, is presented. Once the problem with in-process buffers is represented by the disjunctive graph, it can be solved effectively by using a conventional algorithm. e. g., a branch-and-bound method, for the ordinary job-shop scheduling problem. Second, the problem is described by another type of the MIP and is modeled using an extended timed Petri net which is useful as a tool for various kinds of simulation. Third, the problem is modeled by using a Gantt chart which is often used to represent a schedule graphically. This model is suitable for developing heuristic methods of scheduling.
  • 西松 豊典, 林 一哉, 鳥羽 栄治
    1995 年 31 巻 7 号 p. 941-943
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    It is investigated and discussed how the defects on the outer-ring surface of bearing should be imaged and detected by the computer image processing technique with CCD camera and the Weber's law. After the analogue image of bearing's defect photographed by CCD camera changed the digital image at A/D converter and the binarization was done about the digital image using the Weber's law, the defects on bearing's surface were detected clearly. From the result of single regression analysis between the sensory value about the defect's grade by inspector and the measured defect's area was examined, there was a good agreement between the values calculated by this regression equation and the sensory value by inspectors.
  • 中村 誠, 藤本 正克, 酒井 義郎
    1995 年 31 巻 7 号 p. 944-946
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    Proposed in this paper is a control method based on skilled worker's performance during drying in fish-drying process, which utilizes prediction of the state of material at the end of drying and regulation of drying conditions to attain the prescribed goal before drying. Experimental results showed the usefulness and the applicability of the present method.
  • Ya Jie TIAN, Nobuo SANNOMIYA, Hiroshi NAKAMINE
    1995 年 31 巻 7 号 p. 947-949
    発行日: 1995/07/31
    公開日: 2009/03/27
    ジャーナル フリー
    An observation data is obtained for an irregular swimming trajectory of fish school in a water tank experiment. The motion of fish school under the influence of disturbances is described by a mathematical model with random variables which are modelled as independent, white and Gaussian quantities. A simulation is carried out corresponding to the observation data, and quantitative similarity is obtained in an average sense.
feedback
Top