All issues
Volume 40, Issue 3
Displaying 1-14 of 14 articles from this issue
- |<
- <
- 1
- >
- >|
-
Masahiro ISHIBASHI2004 Volume 40 Issue 3 Pages 279-286
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSCritical flowrate characteristics of very short and step nozzles were investigated experimentally. All the nozzles have identical geometry for their inlet contractions complying with ISO 9300, however, they were cut or have a step exactly at or slightly after the throat. Calibration of these nozzles at various pressure ratios against another critical nozzle connected in series revealed considerable differences in their critical flowrate characteristics. For example, quadrant nozzle that was cut exactly at its throat has a larger flowrate by 0.2% than that of toroidal throat Venturi nozzles when the pressure ratio is low, however, it decreases when the pressure ratio is increased and finally takes the same value as that of the Venturi nozzles at the pressure ratio of 0.5. Step nozzle, which has a very small step at its throat, has the same flowrate as the quadrant nozzle at low pressure ratio but keeps the same value all the way in the critical condition regardless of the pressure ratio. Flow velocity distributions estimated from recovery temperature distributions measured by a thin thermocouple wire traversed in the flow through these critical nozzles explain each flowrate characteristics very well as well as reveal lots of information on their flow structures, which can not be obtained by the traditional measurements.View full abstractDownload PDF (2864K) -
Yutian ZHU, Ichiro HAGIWARA, Yasuhiro MOHRI, Yoshio AKIMUNE, H.S. TZOU2004 Volume 40 Issue 3 Pages 287-292
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSWhen an electric current is used to heat the shape memory alloy (SMA), a strong one can shorten the phase transformation but may overheat the SMA or it's surroundings. Here a method which ensures no overheating and enables the heating process depend on either the heating current or the heating time is proposed. It bases on the detection of the ending of the phase transformation and can be used together with other control processes concerned with heating to prevent overheat. Noticed when martensitic phase begins to transfer to austenitic phase by heating, the variation of resistance increases; after reaches the maximum, it decreases and reaches the minimum again when it completely transfers to austenitic phase, the beginning and the ending of the phase transformation can be detected by monitoring the variation of the SMA's resistance. Based on this principle, a circuit with resistance feed back is proposed here. In the verifying experiment, the SMA is heated for a short period, then it's resistance is determined and the variation of the resistivity is calculated, repeatedly. When the variation of the resistivity increases from the minimum, it means the beginning of the phase transformation; while it falls down to the minimum again, it means the ending of the phase transformation, then the heating current is cut off. The experimental result shows a favorable agreement with the analytical solution. With this method, regardless of however the ambient temperature, the heat convection and heat radiation change, the complete phase transformation of the SMA can be ensured and the overheating can be prevented.View full abstractDownload PDF (1798K) -
Takao HINAMOTO, Keijiro KAWAI, Masayoshi NAKAMOTO2004 Volume 40 Issue 3 Pages 293-299
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper considers the problem of designing an error-feedback coefficient matrix and a coordinate transformation matrix that minimize the output roundoff noise in a closed-loop system with a state-estimate feedback digital controller implemented by fixed-point arithmetic subject to l2-norm dynamic-range scaling constraints. First, we describe the closed-loop system with a state-estimate feedback digital controller. Next, after incorporating error feedback into the state-estimate feedback digital controllers, a technique is developed for optimally designing the error feedback coefficient matrix and the coordinate transformation matrix in the state-estimate feedback digital controller so as to minimize the output roundoff noise under the scaling constraints. Finally, a numerical example is given to illustrate the utility of the proposed technique.View full abstractDownload PDF (1377K) -
Takashi YAMAZAKI, Shin'ichi SUMI, Tadashi INOUE2004 Volume 40 Issue 3 Pages 300-306
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSDroop in rectangular pulses is a problem in crystron power supplies for linear particle accelerators. Reduction methods by superposing auxiliary pulses have been discussed. Though feedback control of the auxiliary pulse waveforms and of the switching times for synchronizing the supplies seems to be useful, any analog or real-time digital control does not seem applicable because of pulse widths at several micro seconds or less. This paper presents a control scheme that the waveform is observed during a pulse generation and control calculations are done based on the stored waveform during the following interval to manipulate the time and waveform in the next pulse generated. The scheme is well suited for many pulsed power applications, in which the repetiton rate is so low as lk pps or less. The representation of the control becomes extremely simple, as the pulse decays rapidly not affecting the pulses that follow and the samples of waveform can be reduced to a small number of feedback variables representing waveform and time parameters to be controlled.
The feasibility and performance are tested in a experimental study on switching time control of micro-second pulses with a 12 bit 20MS/s A/D converter and a PC compatible. A delay time control based on correlation maximization is introduced, which proved to be effective for removing low-frequency components in jitter and keeping steady-state deviation as low as 0.6ns.View full abstractDownload PDF (1479K) -
Toru NAMERIKAWA, Masayuki FUJITA, Roy S. SMITH, Kenko UCHIDA2004 Volume 40 Issue 3 Pages 307-314
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper deals with the H∞ control attenuating initial-state uncertainties. An H∞ control problem, which treats a mixed attenuation of disturbance and initial-state uncertainty for linear time-invariant systems in the infinite-horizon case, is examined. The mixed attenuation supplies H∞ controls with good transients and assures H∞ controls of robustness against initial-state uncertainty. We derive a necessary and sufficient condition of the generalized mixed attenuation problem. Furthermore we apply this proposed method to a magnetic suspension system, and evaluate attenuation property of the proposed generalized H∞ control approach.View full abstractDownload PDF (2356K) -
Hirotaka OJIMA, Kenji NAGASE, Yoshikazu HAYAKAWA2004 Volume 40 Issue 3 Pages 315-321
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper investigate the properties of the impedance matching controller for the uniformly varying damped mass-spring systems and the resultant closed loop systems by invoking the properties of the algebraic function. The analyticity and the positive real property of the impedance matching controller are first investigated. Next, the internal stability of the feedback interconnection of the uniformly varying damped mass-spring system and the impedance matching controller is studied. Finally, the ladder structure of the closed loop system and the impedance matching controller is investigated.View full abstractDownload PDF (1393K) -
Naoki KUMATA, Eitaro AIYOSHI2004 Volume 40 Issue 3 Pages 322-327
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSChaos generated from nonlinear systems includes many periodic points and the estimation of these points is significant. In this paper, in order to estimate these periodic points, we formulate the optimal initial value decision problem for a nonlinear map at first. In the next, we consider the minimization function in this decision problem as a functional of the initial value, introduce the gradient of the functional by using a variational method, and present the method to solve the optimal initial value decision problem, when the fuction of a nonlinear map is given explicitly. In the third, we concerned with approximation of an unknown nonlinear map by using a neural network, and we estimates periodic points in the nonlinear map by using the neural network model. At last, we present the result of simulation by using the these methods.View full abstractDownload PDF (1082K) -
Qiang WEI, Tetsuo SAWARAGI2004 Volume 40 Issue 3 Pages 328-337
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper presents a study on Multiagent Reinforcement Learning (RL) for cooperating work from a social view to solve the problems of individual agent's incomplete world model, conflict of individual interests and inconstant reinforcement from environment. Through a modeling of four software robot agents' cooperative work to balance a ball on a plate, two social RL approaches-observing reinforcement and vicarious reinforcement-are applied to individual RL agents in a multiagent learning environment. The comparisons between social RL and independent RL are provided in several aspects, and we show that social RL accelerates the convergence of agents' learning efficiency in an organization and also contributes to the learning efficiency of unskilled agents, who are newcomers and join to an existing organization.View full abstractDownload PDF (3322K) -
Yuanjin LING, Kazuhiko SUZUKI, Takehisa KOHDA2004 Volume 40 Issue 3 Pages 338-345
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn chemical and nuclear power plants, the safety policies have been established according to the consideration of multiple independent protection systems. Independent protection layer 4 (IPL4) is defined as safety instru-mentation system (SIS). Focusing on IPL4, this paper studies the optimal inspection policy and quantitative risk assessment for the SIS with self-diagnosis function. Using Markov models considering dangerous failure (DF) and safe failure (SF), availability A, expected number N(T) of spurious trips caused by SF and probability Q of DF are obtained. Moreover, the risk I caused by SIS failure is analyzed mathematically. We discuss A, N(T), Q and I are influenced by diagnostic coverage and periodic proof test interval. Furthermore, to fulfill the tolerable risk and to consider the inspection cost, the optimal risk-based policy is also discussed. Illustrative examples are given.View full abstractDownload PDF (1446K) -
Satoshi ITO, Haruhisa KAWASAKI2004 Volume 40 Issue 3 Pages 346-352
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, we present a control method of upright posture under unknown periodic external forces assuming that its period is only known. In order to maintain the upright posture in unknown environment, the information of the ground reaction forces are essential. However, being exposed to the stationary environment providing the external force periodically, the regularity in it is learned, and the unknown factors are clarified using adaptation dynamics. Consequently, the balance maintenance is achieved without feedback of ground reaction forces, and the torque pattern is internally generated to control the balance in the feedforward manner.View full abstractDownload PDF (2471K) -
Toshiyuki KONDO, Norihiko ITOH, Koji ITO2004 Volume 40 Issue 3 Pages 353-360
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn the last decade, a number of skillful robots have been developed. However most of them can so far only demonstrate pre-programmed motions according to the external stimuli. In contrast, humans can learn new motions such as catching a ball in spite of his/her high dimensional sensorimotor DOF. In the learning process, it can be hypothesized that the learner actively constrains the DOF by him/her-self using learning skills, in this paper referred to as “schema”. In this study, a learning method for autonomous mobile robots operating in unknown environments is proposed, where not only a behavior learning mechanism for sensorimotor mappings but also an extraction/re-use mechanism of the schema (i.e. constraint rules for behavior learning) are implemented. Through the results of simulations and real experiments of mobile robot navigation, the validity of the proposed method is clarified.View full abstractDownload PDF (4410K) -
Tomohiro KUBO2004 Volume 40 Issue 3 Pages 361-363
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this short note, the optimal memoryless regulators for systems with time-delay of retarded and neutral type are considered, and it is shown that the poles are related to the hamiltonian as well as the finite dimensional LQ regulator.View full abstractDownload PDF (351K) -
Masanori HIGUCHI, Toshiyuki TANAKA2004 Volume 40 Issue 3 Pages 364-366
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper deals with biometric identification using phase only correlation from hand vein patterns. We made an equipment for image input, which consisted of CCD camera and infrared LED. We obtained 86% certification ratio using our proposed method.View full abstractDownload PDF (602K) -
Takabumi FUKUDA, Toru INOKUBO, Hisaji SHIMIZU2004 Volume 40 Issue 3 Pages 367-369
Published: March 31, 2004
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSTo know the trend of wind velocity hidden in noisy fluctuation, a method using Kalman filters is proposed. The trend is judged probabilistically by comparing the consistence probabilities of several trend models. Through some experiments and experiences, the feasibility will be shown.View full abstractDownload PDF (1138K)
- |<
- <
- 1
- >
- >|