駆動系に存在する摩擦トルクは非線形制御で補償し, 分布系という特性から生じるスピルオーバ現象はH_∞制御による周波数整形で抑制し, 数値的制御系設計法の有力な方法の1つであるQパラメータ表示を適用することによって, H_∞コントローラを適用したときに生じる大きなオーバシュートを改善できることが実験により確認された.さらに, アーム先端に未知質量を有する弾性アームに関してもVSS型適応同定と適応オブザーバにより位置決め制御が可能であることをすでに明らかにしている.このように, より高度な弾性アームの制御が可能となってきている.

抄録全体を表示

本研究は, マスタマニピュレータの製作, バイラテラル制御系の設計と位置決め制御系の改善を行い, その有効性について検証し, 次の結果が得られた.
(1) コンピュータからの直接入力に対する位置決め精度は, 分解加速度制御を用いるのと同程度の結果を最適1型ディジタルサーボ系を用いて得られた.また, バイラテラル制御系でのマスタの軌道へのスレーブの追従精度は分解加速度制御を用いるより最適1型ディジタルサーボ系を用いることで大幅に改善され, 遠隔システムとして十分使用できることが確認できた.
(2) 切断実験では, カセンサが臨場感を得るために有効であることが確認できた.
(3) 剛体接触時に生ずる振動抑制に対しては, 位置決め制御系の改良はあまり有効でないことがわかった.

抄録全体を表示

This paper describes robust cooperative motion control systems based on workspace and joint space disturbance observers for multiple manipulators. In the modeling phase, the workspace disturbance observer based system uses a standard diagonal mass matrix while the joint space disturbance observer based system utilizes a diagonal nominal inertia matrix which results in the acceleration based motion control which means that both approaches provide a simple, flexible and robust cooperative motion control schemes. The effectiveness of the proposed systems is verified by several experimental results using two units SCARA manipulators grasping an object and moving the object according to the desired trajectory in free motion condition as well as in the existence of external interaction force.

抄録全体を表示

Nowadays off-line teaching of a robot is desired to take the place of common manual teaching. It is a precondition for off-line teaching that a robot has high positioning and orientating accuracy. ISO 9283 defines performance criteria of a robot which include these kinds of accuracy and it recommends several systems to test them. These systems are required to measure the position and the orientation of a robot hand in real time, with high accuracy and without contact. As for position measurement, a system using a laser beam has been proposed and used by some users and manufacturers of robots now. However it is very difficult to measure the orientation of a robot and there is no practical measuring system in the present state. In this study an orientation measuring system of a robot using a gyroscope is proposed. Since a practical robot hand performs 3-DOF (degree-of-freedom) complicated rotational motion dynamically, the accuracy of this system under such conditions was inspected by using a precise 3-DOF turntable as a calibration standard. It was proved that the accuracy is about ± 0.03° and ± 0.05° for static and dynamic orientation measurement respectively. It was proved by these results that this system has a good potential to measure the orientation of a practical robot.

抄録全体を表示

Air fluctuation around the fine equipment system seems to be a kind of extremely low velocity air flow and is one of the major causes in deteriorating system working accuracies. Unfortunately there is no active means to control air fluctuation at present. This paper describes the controllability of air flow around the equipment assuming that the air flow dynamics is approximated by a linear equation under the low velocity viscous flow condition. With the verification of its controllability, several experiments are performed to confirm the possibility of air flow control by using smoke particles. Specifically, smoke particles are stimulated by the optical radiation pressure of Ar laser beam. Their behaviors are regulated by the optical system such as a convex lens and an ND filter with the number of stimulated smoke particles. The experimental results basically show the possibility of air flow control using an optical radiation pressure.

抄録全体を表示

In order to maintain the safety of steel tubes which have been used externally as structural members in buildings and large scale structures for a long period, the inner walls of the tubes are often investigated by inspectors. In the present study, to reduce this visual inspection, they have paid attention to the color of the rust which appears in the corroded wall, and have tried to detect corroded regions by using color image processing. Namely, the input RGB color images are transformed to color sensitive space Lab, and three parameters, lightness, chroma, and hue angle are calculated and investigated around the corroded region. Moreover, they have developed this method to be able to inspect the video tape observing corroded regions in real time by using a commercial high speed image processing board.

抄録全体を表示

By using a confocal scanning laser microscope (SLM) which is commercially available, the topography data of surface texture can be easily obtained in a short time. However, unreasonable short-wavelength components tend to be included in the raw data, when compared with the data obtained through a contact stylus instrument. In order to eliminate such noisy components, a digital filtering process was proposed and its design procedure in the spatial frequency domain was developed. The functional performance of the digital filtering was discussed with regard to the surface texturescreated by several machining methods. A fairly good agreement was reached between the corrected SLM data and the reference data in respect of perspective 3-D plots and some conventional roughness parameters. On the other hand, this filtering process showed unsatisfactory effects on certain local geometry parameters of surface texture.

抄録全体を表示

A simple and clear algorithm for dynamic analysis of mechanisms using motor algebra and graph theory has been developed. A lower kinematic pair such as rotational pair, helical pair or translational pair is represented by a pair-axis motor. Higher kinematic pair such as spherical pair or cylindrical pair is represented by a chain of multiple lower pairs having void links between them. Gear pair is represented by a line vector or a screw which is repositioned after every small movement. Cam-follower pair is replaced with an equivalent cross-slider crank chain sized and positioned by the velocity and acceleration functions derived form the cam-diagram. An algorithm for the analysis of jump of cam is also proposed. A closed-loop mechanism and an open-loop mechanism are driven simultaneously to estimate the timing of jump and strike by comparison of acceleration or displacement between the two. In this paper, a new procedure for the small movement is introduced. Transformational matrix at each pair-axis motor is calculated from the relative angular displacement, then superposed along the tree of link-pair graph to determine the absolute displacement of each link. The computational error is reduced significantly by this procedure, which enables the high-speed dynamic simulation.

抄録全体を表示

In Japan, 60% of neonates are born with auricular deformity. Therefore it is important to express the auricular shape numerically and utilize it for medical treatment. In this paper, the auricular shape is approximated by B-spline function of order four with four internal knots. Derived positions of control points are used as an indicator of auricular shape. Neonates' 490 auricular shapes are classified into following six groups : normal ear, upper folded helix, diagonal side folded helix, side folded helix, Stahl's ear and folded lobulus auriculae. Control points of each group are compared and it is revealed that every group can be characterized by those positions. Also discriminant analyses using the position and Mahalanobis generalized distance are performed. For folded lobulus auriculae, upper folded and side folded helix success ratio of more than 0.7 is obtained. However, as diagonal side folded helix and Stahl's ear are unable to distinguish from normal ear sufficiently, those success ratios are less than 0.7. Finally, target auricular shape after treatment is generated by moving control points gradually toward the direction where Mahalanobis distance between normal ear becomes the least.

抄録全体を表示

This paper presents an attempt to automatically and flexibly solve the sequencing problem of machining operations that enable to produce a part under machining constraints. To realize an flexible planning mechanism, two approaches are performed. In these approaches, operation planning problems are modeled as combinatorial optimization problems. First, genetic algorithms (GAs) are applied directly. To implement the GAs, a string is encoded as an order of removing operations. Thus, a process that produces a finished part from a raw material by removing the machinable volumes is represented as a sequence of the numbers of machinable volumes. As the second approach, the classifier system (CS) is applied to this problem to obtain the further advantages. In this approach, a classifier rule corresponds to a production rule that instructs the next removal volume for a current machining state. As a result of the learning, a near optimal process plan is represented as a chain of classifiers with higher strength values. Moreover, a construction of synthesized planning systems based on GA-approach and CS-approach is also discussed. On the basis of the proposed approaches, a genetic based operation planning simulator is constructed and some numerical experiments are carried out.

抄録全体を表示

The author argues that as long as euclidean processing provides the basis for CAD, there will always be a limit to the level of reliability achievable in a system, i.e., a CAD system contains within it elements that inevitably lead to a system's inaccuracy, instability, and complexity because “a division operation is the root cause of all evil”. The author proposes a new geometric processing paradigm in which there is no need of performing a division operation, i.e., “Totally Four-Dimensional Homogeneous Paradigm”. He discusses the paradigm from the points of its theoretical foundations, geometric definitions, geometric computations and topological definitions.

抄録全体を表示

With kinematic analyses of assembly, it is difficult to derive the motion of the assembly because many constraints make the kinematic model of the assembly complex. In order to derive an equivalent minimum kinematic model for specific components, a constraint reduction procedure and reduction rules based on configuration of contact relations are proposed as follows : first, mapping rules of the kinematic model are proposed in order to realize the mapping from a (feature-based) kinematic model to a kinematic model in configuration space. Second, constraint reduction procedure in configuration space is proposed. Third, based on this constraint reduction procedure and the mapping rules, feature-based constraint reduction and reduction rules are proposed by introducing virtual feature. Finally, in order to confirm the usefulness of the proposed method, it is applied to kinematic analysis of assembly.

抄録全体を表示

溝形状が丸で (R溝), 溝深さの異なる種々のラップを組み合わせてラッピング実験を行った結果, 次の結論が得られた.
(a) 研磨速度とウエビネス
溝深さが0.1D～0.3Dの間では, 研磨速度は固定ラップの溝深さにはあまり影響されず, 回転ラップの溝が深いほど大きくなる.研磨速度の大きさと低次のウエビネスの改善との間にはほぼ正の相関があり, 研磨速度が大きいほど低次のウエビネスは減少する.しかし, 中高次のウエビネスに対しては研磨速度が大きいことはよい結果を与えない.
(b) 溝深さとウエビネス
低次のウエビネスは固定, 回転の両方のラップの溝が深い場合 (0.3D+0.3D) に最もよく修正された.浅い溝は開き角の大きいV溝 (平面) と等価な接触状態である.また, 深い溝は開き角の小さいV溝に対応している.
中高次のウエビネスの修正には回転ラップの回転誤差の影響が小さくなるように, 固定側を深く, 回転側を浅くして小さい研磨速度で加工すれば効果がある.

抄録全体を表示

The purpose of the study is to propose new abrasive machining under wet condition and constant pressure by using chaos. Rotation of an abrasive motor is statically controlled to become chaotic. The following facts are revealed by experiments : (1) A period 1.06 s of a chaotic rotational change is superior than other periods in removed volume and abrasive efficiency. (2) Chaotic rotation of a period 1.06 s is twice as good as maximum and constant 12000 rpm in abrasive efficiency. (3) A chaotic rotational change is superior than changes generated by a sine wave or a random signal in both of removed volume and abrasive efficiency. (4) Reduced percentage of final surface roughness is 96.5-99.5 %, when a plane (Ra 2.3-2.7 μm) machined by WEDM is selected as a target. (5) Rotational change is not so serious that the influence in surface roughness is within 0.1-0.2 μm (P-V). (6) The best surface roughness is determined by size of grains and is 50-100 m (P-V), when #120 grains are used under wet condition and constant pressure.

抄録全体を表示

This paper presents a new evaluation method of wettability between solid and liquid metals in ultra-high vacuum. An argon ion sputter cleaning process of the specimen and a highly magnified observation of the contact angles by scanning electron microscope were adopted to obtain the most reliable experimental values of the contact angles. It was actually found that the wettability between silver and tungsten, that had been thought to be poor, was rather good, when both of the surfaces were perfectly cleaned in ultra-high vacuum. The contact angles for various liquid metals on tungsten were also determined using this experimental method. The empirical tight binding calculations were performed and the local density of states was used to evaluate the wetting interaction energy. The calculated local values were qualitatively proportional to the experimental values of the interaction energies. In order to control wettability, it is necessary to control the electronic states of the substrate surface.

抄録全体を表示

Process planning for machined parts typically requires reasoning about machining features. However, such machining features do not generally take into account a micro-level geometric characteristics of contacting shape between the required shape and the cutting tool, and the macro-level geometrical characteristics of removal volume. Therefore these machining features do not correspond directly to machining operations. The objective of this paper is to capture the fundamental characteristics of machining process and make suitable machining process models based on them. First, a model of the machining process based on the form shaping system theory proposed by Reshetov is constructed. Second, the characteristics of machined surface are the number of parameters in the family of the instantaneous positions of the cutting points, the machined surface, and the contact shape between the locus of the cutting edge in a cutting motion (named macro cutting tool) and the machined surface, are proposed. The characteristics of machining process are the number of macro cutting tool parameters, the number of feed directions in machining operation, and the number of enveloping constraints are also proposed. Finally, the abstract model of machining operation and the abstract model of operation planning which can represent the linear equations are proposed.

抄録全体を表示