All issues
Volume 36, Issue 9
Displaying 1-12 of 12 articles from this issue
- |<
- <
- 1
- >
- >|
-
Yimin SHAO, Kikuo NEZU2000 Volume 36 Issue 9 Pages 729-735
Published: September 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe relationship of failure detection ratios between sensors and bearings at different distances were investigated in order to decide the rational numbers of sensors and select appropriate sensor locations. The impulse signals from damaged bearing were separated from the vibration signals of machinery failure by an envelope detector. The failure detection ratios were calculated according to the analysis of impulse signals. Experimental results have shown that locating an accelerometer on the bearing housing to monitor conditions of a few bearings at different distances is possible and the relationship between the failure detection ratios and various testing conditions have been obtained. This detection method, using an accelerometer to monitor a few bearings at different distances, has been proved to be able to satisfy the various requirements of failure diagnosis. In addition, since the method reduced the unnecessary number of sensors and other devices, it can not only decrease the costs of monitoring devices, but also decrease the self-maintenance costs of monitoring devices. This study provides a new approach for bearing failure diagnosis.View full abstractDownload PDF (2737K) -
Yukihiro HONMA, Yuji KUBOTA, Takashi KASUGA, Jun NISHIKAWA2000 Volume 36 Issue 9 Pages 736-742
Published: September 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper describes a fringe tracking system using a new method to detect an optical path length difference (optical delay) between the ray paths from the interferometer element telescopes. This system was developed and tested for a stellar optical interferometer, called Mitaka optical and infrared array (MIRA), in National Astronomical Observatory. With such a system, the center of interfrarence fringes should be tracked with an accuracy of less than a wavelength by detecting and compensating optical delay fluctuation due mainly to atmospheric turbulence in real time. Our fringe tracking system advanced in reliability and efficiency of central fringe tracking, in bandwidth and stability of servo, and in simplification of system, while fringe tracing has been succeeded at three stellar interferometers before. We adopted three wavelength path length modulation, which enables us to compare amplitude of three fringes, measuring fringe phase at all times. As a result, the largest amplitude fringe was always tracked and the time for moving to the central fringe became short in spite of one detector system. We also adopted an algorithm of moving average phase demodulation (MAPD) method to produce an optical delay error signal. A modified digital PID filter was used as a controller of servo. The MAPD method and the PID filter provided stable and fast control against to disturbance. The closed-loop bandwidth of the servo was approximately about 120Hz. The fringe tracking error of 61nm RMS was estimated for the optical delay fluctuation model at a good seeing condition in the summer of Mitaka.View full abstractDownload PDF (1934K) -
Hirohisa SAKURAI2000 Volume 36 Issue 9 Pages 743-748
Published: September 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe triple point of equilibrium hydrogen is realized by an open cell method using a closed-cycle helium refrigerator. Abnormality is observed on the heat capacity curves of solid hydrogen at temperatures of about 0.1K below the triple point of equilibrium hydrogen in the case of usage of 2g of activated ferric hydroxide as a catalyst for the ortho- and para-hydrogen's equilibration. No abnormality is, however, observed in the case of 0.2g of a catalyst against about 0.1mol of hydrogen. In the latter case, it takes about 3 days for equilibration of ortho- and para-hydrogen and flat melting curves are obtained.View full abstractDownload PDF (2460K) -
Kenji FUJIMOTO, Toshiharu SUGIE2000 Volume 36 Issue 9 Pages 749-756
Published: September 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper is concerned with the asymptotic stabilization of a class of nonholonomic systems which are in generalized Hamiltonian formulae. First, it is shown that such systems can be transformed into canonical forms which have specified structure matrices via generalized canonical transformations. Second we will clarify a construction of non-smooth Hamiltonian functions which render the feedback system asymptotically stable. Further a concrete procedure to design a stabilizing controller based on non-smooth Hamiltonians will be derived, and some numerical examples are given as well to show the effectiveness of our technique.View full abstractDownload PDF (3668K) -
Ichijyo HODAKA, Masayuki SUZUKI2000 Volume 36 Issue 9 Pages 757-765
Published: September 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn many literature, it has been shown that stabilizability/robust stability of singularly perturbed systems are equivalent to stabilizability/robust stability of their reduced and boundary layer systems. This paper shows that these facts can be explained by the fact that two steps of the decomposition of full-systems and feedback operation are completely interchangeable. It is shown that the interchangeability enables us to obtain simple viewpoints and general results for stabilization and robust stability analysis of singularly perturbed systems.View full abstractDownload PDF (1877K) -
Hiroshi FUJIMOTO, Yoichi HORI, Atsuo KAWAMURA2000 Volume 36 Issue 9 Pages 766-772
Published: September 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, a novel perfect tracking control method based on the multirate sampling control is proposed, in which the concept of the two-degree-of-freedom preview control is employed. In the proposed method, it is assumed that the usual single-rate robust feedback controller such as the disturbance observer or the H∞ controller already exists, and only the feedforward controller is designed by using the multirate sampling control. The advantages of the proposed method are that 1) the controller can be designed without considering the unstable zeros of the discrete-time plant, 2) the states of the plant match the desired trajectories at every sampling point, 3) high robust performance is assured by the robust feedback controller.
An illustrative example of position control using a servo motor is presented, and the advantages of this approach are demonstrated.View full abstractDownload PDF (1624K) -
Haruhisa KAWASAKI, Ryo TANIUCHI2000 Volume 36 Issue 9 Pages 773-779
Published: September 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper presents an adaptive force-position control in task coordinates taking constraint state transitions of end-effector into account for constrained robot manipulators. In general, the constraints can be represented separately by the subset of generalized coordinates in task space. This controller works on a concept of orthogonalization between force and motion in the subspaces derived form constraints. The gains can be selected independently in each subspace. The friction force at contact point is also taken into account. The asymptotic convergence for both force and motion tracking errors is guaranteed by the Lyapunov-Like Lemma. Experimental results for a three D.O.F robot are shown to demonstrate the effectiveness of proposed method.View full abstractDownload PDF (1375K) -
Yoshito OHTA2000 Volume 36 Issue 9 Pages 780-788
Published: September 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper derives Hamiltonian formulas for singular values of Hankel operators and mixed Hankel-Toeplitz operators. Given an inner function and a continuous time linear system, it is shown that there is a discrete-time linear system whose input-output space is the left shift invariant subspace associated to the inner function. Using this lifting technique, a Schmidt pair of the operators are characterized as a solution to a discrete-time two-point boundary value problem. This results in a transcendental equation for singular values.View full abstractDownload PDF (2002K) -
Lien T.H. BUI, Atsuko KASHIWAZAKI, Yoshihisa TAKAHASHI, Hitoshi KANOH2000 Volume 36 Issue 9 Pages 789-796
Published: September 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSWhen using car navigation systems, it is important for users to be able to find alternative routes in real time when traffic conditions change as well as optimal routes before starting a journey. In this paper, we propose a new method for determining the easiest to-drive and quasi-shortest route within a given amount of time using a genetic algorithm. Genetic algorithms always have solutions in a population to provide alternative routes using other solutions. Parts of arterial roads are regarded as viruses, and we generated a population of viruses in addition to a population of routes. Experimental results show that the proposed method results in a faster search time than the Dijkstra algorithm.View full abstractDownload PDF (2800K) -
Manabu KOTANI, Takeo MIYATA, Seiichi OZAWA, Kenzo AKAZAWA2000 Volume 36 Issue 9 Pages 797-803
Published: September 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe detection of gas leakage sound from pipes is important in petroleum refining plants and chemical plants, as often the gas used in these plants are flammable or poisonous. In order to establish the acoustic diagnosis technique for the leakage sound, we examined the application of modular neural networks to the stable detection. The modular neural network has the ability to adapt its structure according to the environment. Experiments were performed for an artificial gas leakage device with various experimental conditions to imitate the change of environment for a long term. The discrimination accuracy with the proposed network was observed to be about 93%. From the results, we confirmed the effectiveness for the application of the modular neural network to the detection of the leakage sound for the practical use.View full abstractDownload PDF (2548K) -
Yeboon YUN, Hirotaka NAKAYAMA, Tetsuzo TANINO2000 Volume 36 Issue 9 Pages 804-809
Published: September 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn DEA (Data Envelopment Analysis), CCR model, BCC model and FDH model are known as representative models incorporating preference structure of decision makers. These DEA models differ in production possibility set when considering dual problems. In this paper, we formulate dual problem (GDEAD) to generalized data envelopment analysis (GDEA) and define the concept of “αD-efficiency” for the problem (GDEAD). Furthermore, we establish theoretical properties on relationships between GDEAD, model and existing DEA models. Finally, through a numerical example, we show a dominant relation among decision making units with varying α in the problem (GDEAD) and interpret an implication of the optimal solution to the problem (GDEAD).View full abstractDownload PDF (2031K) -
Hiroshige INAZUMI, Narihiro OMOTO2000 Volume 36 Issue 9 Pages 810-817
Published: September 30, 2000
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn the large-scale rule-based systems, it is assumed that addition, deletion, or changes of rules are frequently carried out. Thus, various contradiction or fatal errors would be caused, even if only a few modifications are carried out, where such kind of errors requires to be verified for maintaining the right reasoning. In this paper, we propose a new scheme for verifying rule-based systems using Petri nets. Especially, considering the property of Transition-invariant (T-invariant), the simple method of detecting structural errors in rule-based systems is clarified. That is to say, some reasoning processes are regarded as various series of rules, which are represented by a set of multiple vectors of T-invariant. Using these vectors, the algorithm for detecting what kind of errors exists in the reasoning processes is derived, where the structural errors of the rule set considering here are the series of rules which should not exist on the rule set, and are indispensable to be searched, detected and corrected.View full abstractDownload PDF (1791K)
- |<
- <
- 1
- >
- >|