All issues
Volume 43, Issue 1
Displaying 1-10 of 10 articles from this issue
- |<
- <
- 1
- >
- >|
-
Tsutomu FUKAURA, Kimio SASAKI2007 Volume 43 Issue 1 Pages 1-8
Published: January 31, 2007
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSThis article proposes a simplifid version of previously proposed simultaneous estimation of3D-position and velocity of a moving stationary random acoustic source in near-field by correlation analysis of wavelet patterns of 4 detected signals. By using a simplified wavelet transform at a certain fixed scale value, the method uses 1D-correlation analysis of the wavelet patterns in place of the2D-one in the previous, so that it may save much computation time. The effectiveness of the proposed method is made clear by the results of computer simulations.View full abstractDownload PDF (3940K) -
Tokihiko KOBATA2007 Volume 43 Issue 1 Pages 9-16
Published: January 31, 2007
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSTo improve consistency between gas and hydraulic pressure standards, two methods for comparing pressures of different mediums generated by gas and hydraulic pressure balances are examined. One is the method to use a gas-liquid exchanger and another is the method to use a liquid-lubricated pressure balance. The pressure range compared is0.5MPa to7MPa. In either method, a precise pressure transducer was used as a comparator to determine the equilibrium state between two pressure balances. In this paper, the comparison results and their uncertainties obtained from the respective methods are presented. From the results, it was shown that the results obtained from both methods were agreed within the combined standard uncertainty in all the pressure range compared and the relative difference was sufficiently small compared with the uncertainties obtained from the respective methods. Therefore, it can be stated that both methods adopted in this study are reliable to comparegas and hydraulic pressure balances precisely.View full abstractDownload PDF (4235K) -
Shinichi YAMAKAWA, Kenji AMAYA, M. PARAMESWARAN (ASH)2007 Volume 43 Issue 1 Pages 17-22
Published: January 31, 2007
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSThis paper presents a nano-resolution measurement of magnetic fields near a magnetic probe using a thin-film magnetic sensor. The measurement method consists of a process, where the thin-film magnetic sensor is scanned over the sample, followed by a deconvolution scheme. In the scanning process, the thin-film magnetic sensor is scanned over the sample perpendicularly to the sensor. The sample is rotated horizontally under the sensor so that data measurements are obtained at different angles and positions. The deconvolution can be performed using existing methods, which are common in the field of computed tomography. An experiment was performed to verify the technique's capability of measuring a magnetic field using the thin-film magnetic sensor. Magnetic force microscopes are widely used for measuring magnetic fields. However, the magnetic field from the magnetic probe affects the magnetization of the sample when the sample's magnetization is small. Thin-film magnetic sensors do not have a magnetic field. Therefore the presented technique can measure small magnetic fields even from very sensitive magnetic samples. Currently, thin-film magnetic sensors are used as reading heads in hard drive disks (HDD). Types of heads used are thin-film inductive heads, anisotropic magnetoresistive (AMR) heads, and giant magnetoresistive (GMR) heads. Nowadays, the film thickness of the GMR head, which is most commonly used as the reading head in HDD, is less than 10nm. Previously, the resolution of the measurement was limited by the width of the sensor (ie. 100nm) which is always greater than the thickness. However, the new technique has the same resolution as the film thickness (ie. 10nm). The measurement techniques used in the HDD industry can be applied to our measurement method.View full abstractDownload PDF (10785K) -
Daisuke AKASAKA, Kang-Zhi Liu2007 Volume 43 Issue 1 Pages 23-30
Published: January 31, 2007
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSThis paper studies the feedback control of a class of second order linear systems which include the double integrator and are actuated by on-off input. This class of systems includes many important applications such as the orientation control of satellites which are actuated by thrustors. A class of feedback control laws are proposed which guarantee the global stability. The basic approach in stability analysis is to use continuous and piecewise smooth Lyapunov functions and Filippov theory.View full abstractDownload PDF (3716K) -
Masato ISHIKAWA, Ichiro MARUTA, Toshiharu SUGIE2007 Volume 43 Issue 1 Pages 31-36
Published: January 31, 2007
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSThis paper deals with discrete-valued controller design for linear systems considering its limitations on temporal and spatial resolutions. First, we introduce a feedback modulator (FBM), which converts continuous-valued control input to discrete-valued one. It is composed of a feedback connection of sampler, holder, quantizer and a stable filter as a tunable parameter. Then it is combined with a continuous-valued compensator appropriately designed in advance. Based on the analysis of performance and stability issues for the FBM and FBM-based control systems, we provide a sufficient condition for the tunable filter to stabilize the whole system, in terms of relation between its time constant and the temporal resolution. This condition gives as an easy-to-follow guide to design a practical FBM. Effectiveness of the proposed method is validated with numerical simulations, with comparison to conventional approaches.View full abstractDownload PDF (3716K) -
An Approach to Inverse Problems with the Two-sided z TransfornTakuya SOGO2007 Volume 43 Issue 1 Pages 37-45
Published: January 31, 2007
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSSince the transfer function of sampled-data systems with a short sampling period frequently has zeros outside the unit circle, it is recognized that iterative learning control (ILC), which is an iterative approach to obtain an inversion of a system, cannot be applied to such systems. In order to overcome this difficulty, we introduce a non-causal bounded inversion to a sampled-data system and investigate the limiting property of the inversion. We demonstrate a simple relation between the η-th differential operator and n Euler operators that are included in the pulse transfer function of the η-th integral element. Based on this relation, we prove that the inversion of the sampled-data system approximates its continuous-time counterpart in a non-causal framework. This result suggests that ILC can be applied to sampled-data systems in this framework.View full abstractDownload PDF (4716K) -
Sagittal Motion Control and StabilizationMasahiro DOI, Yasuhisa HASEGAWA, Toshio FUKUDA2007 Volume 43 Issue 1 Pages 46-53
Published: January 31, 2007
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSIn this paper, Passive Dynamic Autonomous Control (PDAC), the novel control method we proposed previous paper, is applied to the sagittal motion of biped walking. PDAC expresses the robot dynamics as an 1-dimensional autonomous system based on the two concepts: 1) point-contact 2) Virtual constraint (proposed by Grizzle and Westervelt et al. previously). Due to autonomy, this approach has two following notable point: 1) there is a conservative quantity2) period from foot-contact to next foot-contact can be obtained by integrating1D dynamics with respect to time. This paper proves the convergence of the conservative quantity under the condition of constant step-length by Theorem of Lyapunov. In addition, the coupling method of the sagittal and lateral motion that takes advantage of point2) is designed. Finally, by means of PDAC and these methods, natural3D dynamic walking based on the robot inherent dynamics is realized.View full abstractDownload PDF (12364K) -
Katsunari SHIBATA, Koji ITO2007 Volume 43 Issue 1 Pages 54-63
Published: January 31, 2007
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSHumans can learn its action effectively in the real world by taking a similar action in a similar situation. In order to realize such abilities in a robot with a neural network, not only generalization on the input signal space, but also the generalization based on the similarity on the reconstructed space on the hidden layer would play an important role. In this paper, to explain the acquisition of the useful hidden representation for the generalization, a hypothesis is set up at first that when two training patterns are closer, the corresponding hidden patterns are likely to become closer through learning. In the first simulation, the hypothesis is supported by the learning of random input-output pattern sets. In the other simulations where simple visual sensor signals were the input of the network, it is shown that the hidden layer represents global information adaptively while keeping the information given by the initial connection weights from the input layer to the hidden layer. Finally, the reason why a neural network with visual inputs becomes to represent global information in the hidden layer through the reinforcement learning of the task in which a robot reaches a target caught on its visual sensor, is considered.View full abstractDownload PDF (16195K) -
Wataru KASE2007 Volume 43 Issue 1 Pages 64-66
Published: January 31, 2007
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSRecently, a complete general parametric solution to the generalized second-order Sylvester matrix equation was presented. In this note, it will be shown that the idea is also valid for an arbitrary order. It will be also shown that the calculation of the solution can be easily obtained by using the basis in polynomial vector space.View full abstractDownload PDF (325K) -
Hiroaki MUKAIDANI, Seiji YAMAMOTO, Hua XU2007 Volume 43 Issue 1 Pages 67-69
Published: January 31, 2007
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSIn this paper, a numerical algorithm for solving the H2/H∞ control problem is investigated. Although the equations consist of the nonlinear cross-coupled algebraic Riccati equations, it is shown for the first time that even if γ is relatively small, the local uniqueness and the quadratic convergence for the iterative solutions are both guaranteed.View full abstractDownload PDF (472K)
- |<
- <
- 1
- >
- >|