All issues
Volume 41, Issue 3
Displaying 1-13 of 13 articles from this issue
- |<
- <
- 1
- >
- >|
-
Jun TAMBA, Koichi SATO, Kazuaki YAMAZAWA, Isao KISHIMOTO, Masaru ARAI2005 Volume 41 Issue 3 Pages 187-192
Published: March 31, 2005
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA new apparatus consisting of a fixed point cell and a furnace was developed to evolve the accuracy of realizing the freezing point of tin (231.928°C) defined in the International Temperature Scale of 1990 (ITS-90). The temperature stability and uniformity of the furnace are better than a few millikelvins. This apparatus can realize the freezing plateaux lasting longer than 45 hours, with a slight temperature drop within 0.3mK. From this prolonged phase transition, the freezing process will come closer to the ideal freezing point under the thermodynamic equilibrium. Since the tin point shows a large supercooling in the freezing process compared to the other freezing points, a newly designed gas-cooling instrument was introduced for safe operation and precise cooling control. Such the apparatus and the optimized procedure give satisfactory repeatability within 0.026mK from ten realizations of the freezing point. These results confirm the capability of the new apparatus for replacing the current one as the national standard.View full abstractDownload PDF (2614K) -
Yuichi IKEDA, Ken-ichi AOTO, Takashi KIDA, Tomoyuki NAGASHIO2005 Volume 41 Issue 3 Pages 193-201
Published: March 31, 2005
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, we consider six dof (degree of freedom) control of rigid body spacecraft via nonlinear H∞ control method. Firstly, the motion equation of the six dof rigid body spacecraft is described and the error system is derived. Secondly, we show that the error system can be stabilized by PD control law with the constant coefficient as the case of attitude control and that it guarantees structural robust stability with respect to physical parameter error. Thirdly, based on nonlinear. H∞ control, the design conditions of a gain that the control law becomes suboptimal H∞ state feedback control are given. Finally, effectiveness of the control methods is verified by simulations.View full abstractDownload PDF (1465K) -
Control System Design Based on the Principle of MatchingTakahiko ONO, Hikaru INOOKA2005 Volume 41 Issue 3 Pages 202-209
Published: March 31, 2005
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSWhen an ambulance changes a lane or turns a corner, the lateral acceleration acts on a patient. Such acceleration causes patient's discomfort, carsickness and physical pain. To reduce the effect of lateral acceleration on a patient, we made the prototype of an actively-controlled bed for ambulances that rotates the bed around the axis of a patient's body with a DC servo motor. The purpose of this paper is to report the control system design for this actively-controlled bed and to evaluate its control performance in simulation and experiment. In the framework of the principle of matching, the control system is designed as a servo system so that the tracking error and the motor torque are maintained within the tolerable bounds in actual situation. The design problem is formulated to a feasibility problem stated as inequality constraints. A numerical search is performed to find parameters of a controller. In experiment, it is confirmed that the lateral acceleration acting on a patient is reduced to about 25% as compared to use of normal fixed beds. In simulation, the cancellation rate of lateral acceleration is examined when the rotational angle of the bed is restricted to ±5, ±10 or ±15 degrees.View full abstractDownload PDF (2928K) -
Kazuo KOMATSU, Hitoshi TAKATA2005 Volume 41 Issue 3 Pages 210-215
Published: March 31, 2005
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, we deal with a formal linearization algorithm of a general class of time-variant nonlinear systems using interpolation of orthogonal polynomials. A linearizing function which consists of the Chebyshev polynomials is defined. The time-variant nonlinear systems are transformed into time-variant linear ones with respect to this linearizing function by exploiting Chebyshev interpolation to the state variable and Laguerre interpolation to the time variable. This kind of linearization has simple algorithms due to the orthogonality for a finite sum, so they are easily executed by using computers. The inversion of this linearization is also simple. Besides, a time-variant nonlinear observer algorithm is presented as an application of the linearization.View full abstractDownload PDF (1330K) -
Rong HE, Kang-Zhi LIU2005 Volume 41 Issue 3 Pages 216-225
Published: March 31, 2005
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA nonlinear dynamic output feedback control method for a 5-degree-of-freedom active magnetic bearing system is presented in this paper. This method guarantees the asymptotic stability of the closed loop system with arbitrarily low bias current. The basic idea is to make full use of the feature that the dynamics part is linear and hence can be stabilized by output feedback when the output of actuator is regarded as the virtual input of the linear dynamics. Then, a quadratic-like Lyapunov function is proposed for the nonlinear design of control voltage and it is shown that backstepping and completing square techniques enable the construction of the voltage input by using the measured output only.View full abstractDownload PDF (1721K) -
Takahiro ENDO, Fumitoshi MATSUNO2005 Volume 41 Issue 3 Pages 226-233
Published: March 31, 2005
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSWe discuss the controller design and stability of closed-loop system for two-link flexible arm. The flexible arm is modeled as distributed parameter systems, which is infinite dimensional. In general, a controller for such distributed parameter systems is designed based on a finite dimensional approximated model. Therefore we should consider spillover instability that is the drawback resulting from finite dimensional approximation. In this paper, as we design the controller based on an original distributed parameter model, we can avoid the drawback. At first, using Lyapunov method, we derive a PDS feedback controller that the closed-loop system becomes Lyapunov stable. Note that the proposed PDS controller differs from the conventional PDS controller. The conventional PDS controller works well in neighborhood of a desired configuration of the arm, but the pro-posed PDS controller can ensure global stability. Next, we prove that the proposed PDS controller can ensure the asymptotic stability of the closed-loop system at the neighborhood of the desired configuration of the arm. The proof is completed using LaSalle's invariance principle, which is extended to infinite dimensional systems. Finally, in order to demonstrate the validity of the proposed controller, experiments have been carried out.View full abstractDownload PDF (2582K) -
Yoshiaki ASAGI, Keiji WATANABE, Eiichi MURAMATSU, Yuichi ARIGA2005 Volume 41 Issue 3 Pages 234-241
Published: March 31, 2005
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper presents a systematic design method of internal stable decoupling of multivariable systems on state space. First, internal stable decoupling by state feedback is presented for systems with unstable row zeros. Second, unstable zeros which are not row zeros are considered and the internal stable decoupling method is proposed by combining the above method and the compensator. Third, these decoupling methods are applied to decoupling by inverse system. In particular, the inverse system for unstable non-row zeros is presented by modifying the compensator.View full abstractDownload PDF (1630K) -
Manabu KOSAKA, Hiroshi UDA, Eiichi BAMBA, Hiroshi SHIBATA2005 Volume 41 Issue 3 Pages 242-249
Published: March 31, 2005
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, we propose a new deterministic off-line identification method that delivers a state-space model by using input and output data with steady state values. This method combines the method zeroing the 0N-tuple integral values of output error of single-input single-output transfer function model and Ho-Kalman's method. The feature is that this method can identify by using step responses and, consequently, this is suitable for a mechanical system identification in which persistent fluctuations are unacceptable. Numerical simulations of multi-input multi-output system identification are illustrated and the practicability is verified by application to a pneumatic two link arm system.View full abstractDownload PDF (1697K) -
Nami KIDANE, Hisakazu NAKAMURA, Yuh YAMASHITA, Hirokazu NISHITANI2005 Volume 41 Issue 3 Pages 250-258
Published: March 31, 2005
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, we propose a two-step controller design method for a nonlinear system with an input constraint by using a local control Lyapunov function. Malisoff et al. showed a universal control formula for a nonlinear system such that the k-norm of inputs is less than one, where k is limited to 1<k≤2. First, we propose a new control formula that can be applied in any case of k≥1. The inputs are continuous at any points except the origin in the case of 1<k<∞, but these inputs may become discontinuous in the case of k=1 or k=∞. This causes a chattering phenomenon. Second, we improve the control formula in order to make inputs continuous in any case of k≥1. Finally, we show the effectiveness of this control formula by computer simulation.View full abstractDownload PDF (2087K) -
Yuanjin LING, Kazuhiko SUZUKI, Takehisa KOHDA2005 Volume 41 Issue 3 Pages 259-267
Published: March 31, 2005
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn chemical and nuclear power plants, protective systems are used for risk reduction. Now, with rapid development of microprocessor technology, the protective system with self-diagnosis function has spread. This paper studies the effect of self-diagnosis on protective systems consisting of k-out-of-n: G systems. Considering selfdiagnosis, scheduled and unscheduled maintenances, expected numbers of normal trips, failure detection trips, spurious trips and hazards caused by protective system failure are derived. We discuss how the variation of expected number of failures are influenced by the failure detection ratio. Moreover, the total expected loss caused by protective system failures is analyzed mathematically as the sum of unscheduled maintenance cost and loss cost. The optimal failure detection ratio that can minimize the total expected loss is also discussed. Illustrative examples are given.View full abstractDownload PDF (1765K) -
Kazuaki MASUDA, Eitaro AIYOSHI2005 Volume 41 Issue 3 Pages 268-273
Published: March 31, 2005
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSPrincipal component analysis (PCA) is a fundamental tool for processing multidimensional data. Although various algorithms for finding principal directions have been proposed, they are interpreted as optimizing a certain objective function with orthogonal constraints which is incorporated into its Lagrangean as penalty functions.
In this papar, we propose a new method based on constrained optimization technique which directly transform the constrained decision variables into the unconstrained ones through nonlinear functions. This method takes advantage in numerical computations and has extensibility to the global optimization aided by chaos. We start from reformulating PCA as an optimization problem on a hyper-spherical surface with orthogonality constraints. Then, we propose a new nonlinear transformation by considering the rotational operation in Euclid's space. Finally, we derive its equivalent unconstrained optimization problem and provide the gradient dynamics for solving it analytically. The validity of our proposing methods is demonstrated with numerical simulations.View full abstractDownload PDF (1582K) -
Tetsuya YAKAWA, Hitoshi IIMA, Nobuo SANNOMIYA2005 Volume 41 Issue 3 Pages 274-282
Published: March 31, 2005
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe one-dimensional bin packing problem is one of typical combinatorial optimization problems. This problem has been solved so far by using several metaheuristic methods such as grouping genetic algorithm, variable neighborhood search method and perturbation MBS' method, and by using branch-and-bound based methods such as bin packing solution procedure. In this paper, a design of the genetic algorithm is proposed for the bin packing problem in such a way that diverse offsprings are generated. The crossover operation is designed so as to inherit the combination of items in each parent's bin. The mutation operation is designed by embedding a heuristic rule in such a way that items fit the capacity constraint as much as possible. It is observed from computational experiments for many benchmarks of this problem that the proposed genetic algorithm finds optimal solutions most frequently among other metaheuristic methods.View full abstractDownload PDF (1685K) -
Masaru OKAMOTO, Nan BU, Toshio TSUJI2005 Volume 41 Issue 3 Pages 283-290
Published: March 31, 2005
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper proposes a novel hierarchical clustering method. The radical distinction from traditional methods is that the proposed method requires no specific knowledge of the number of classes for classification. In this method, at each node of a herarchical classification tree, a log-linearized Gaussian mixture network is utilized for clustering, and a newly invented learning law is applied to train the LLGMN unsupervisely to classify data into two subclasses based on statistical characteristics. This method performs a binary classification hierarchically and, finally conducts a classification with a suitable number of classes. Also, unnecessary structure of the classification tree can be avoided using cross-validation. Validity of the proposed method is demonstrated with classification experiments on artificial data and electromyogram (EMG) signals.View full abstractDownload PDF (1939K)
- |<
- <
- 1
- >
- >|