All issues
Volume 39, Issue 12
Displaying 1-14 of 14 articles from this issue
- |<
- <
- 1
- >
- >|
-
Kozo OHTANI, Mitsuru BABA2003 Volume 39 Issue 12 Pages 1075-1084
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA proposed system consists of a sensor array unit, a signal processing unit, and a neural network unit. The sensor array unit acquires the intensity of an ultrasonic irradiated from transmitter to an object. The signal processing unit picks up a necessary data from the output of the sensor array unit. We used the following parameters for the recognition of shapes and for measurement of both positions and poses of the measured objects; the peak positions, peak values, the value of the inflection points and minimum values of the ultrasound pressure distribution. These values and positions are inputted into the neural network unit for object identification, postion and pose measurement. We designed and built a prototype system, and we tested this system by attempting to identify six objects. The experimental results showed near-perfect object identification and high resolution measurement.View full abstractDownload PDF (5160K) -
A Study on Posture Sensor which combined Square Pole Silicon Electrode with Platinum Point ElectrodeYoshito SHIRAI, Yoshifumi KITAYAMA, Keishi KAWABATA, Takeshi TANAKA2003 Volume 39 Issue 12 Pages 1085-1092
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn the paper, we will propose some new idea regarding posture sensor.
This proposal is a measurement technique of the posture by combining the detection method of discrete liquid level by the point Pt electrode with the liquid level measurement method by square pole Si electrode that has already been proposed by authors'. Using these techniques, we succeeded in improving the precision and expansion of the range of the measurement.
First of all, we will explain the simulation technique to examine the mounting of the square pole Si electrode. Next, we will show calibration experiment method of square pole Si electrode, posture measurement, temperature dependency experiment, and the addition function of the attitude measuring method that uses liquid film behavior.
As a result of these experiments, we came to the conclusion that our proposal has validity.View full abstractDownload PDF (2644K) -
Shin KANNO, Gan CHEN, Hiroshi SHIBATA, Toru FUJINAKA2003 Volume 39 Issue 12 Pages 1093-1098
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper considers a problem of stabilization for input delayed systems. Using redundant descriptor representation, we derive a sufficient condition for stabilization of input delayed systems via static state feedback. Furthermore, we give a robust stabilization condition for input delayed systems whose coefficient matrices depend rationally on uncertainty parameters. These conditions are delay-dependent solutions and are derived by computable finite set of linear matrix inequalities. Finally, an illustrative numerical example is given.View full abstractDownload PDF (1350K) -
Kiyoshi YOKOI, N.V.Q. HUNG, H.D. TUAN, Shigeyuki HOSOE2003 Volume 39 Issue 12 Pages 1099-1107
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA novel adaptive backstepping design for a class of n-th order nonlinearly parameterized systems with a triangular structure is proposed. Under the Lipschitz condition with respect to unknown parameters of the system, an effective adaptive controller is designed without the requirement on the compactness of the unknown parameter set. Especially, the proposed adaptive control enables the advantage of “tuning function concept”, which results in only one estimation law for the unknown parameters. Our simulation with induction motor model particularly shows the viability of the obtained results.View full abstractDownload PDF (2109K) -
Hiroaki YAMAGUCHI2003 Volume 39 Issue 12 Pages 1108-1116
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper presents a distributed control scheme for multiple nonholonomic wheeled mobile robots of the Hilare-type to make group formations. The Hilare-type mobile robot is kinematically equivalent to a unicycle-type mobile robot. Each robot in this control scheme has its own coordinate system, and it senses and controls its relative position to other robots, which means that each robot has relative position feedback. Via this feedback, this robotic system consisting of the multiple mobile robots becomes a large-scale visually articulated multi-body system. Each robot especially has a two-dimensional control input referred to as a “formation vector” and the formation is controllable by the vectors. The mobile robots have nonholonomic constraints and they cannot move in an omni-direction instantaneously, so that such nonholonomic mobile robots cannot be asymptotically stabilized by smooth static feedback control laws. We introduce a distributed smooth time-varying feedback control law whose asymptotic stability is guaranteed in a mathematical framework, averaging theory. To prove its asymptotic stability, we apply Eckhaus'/Sanchez-Paleucia's theorem. We specifically deal with a matrix whose components are averaged over time by integration and whose eigenvalue distribution describes stability of this control law. Since the eigenvalue distribution is analytically shown by Hölder's inequality in functional analysis, stability analysis of this control law is given in averaging theory and it is technically related with functional analysis. The validity of this control law is supported by computer simulations.View full abstractDownload PDF (4519K) -
Toshiyuki OHTSUKA2003 Volume 39 Issue 12 Pages 1117-1123
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper considers quadratic-in-the-state representations, which consist of state equations that are at most quadratic with respect to the states, as representations for a broad class of nonlinear systems. A necessary and sufficient condition is shown for existence of a quadratic-in-the-state representation that has the identical input-output map with a given nonlinear system. That condition is characterized by the algebraic structure of the observation space of the given system and is so mild that many types of nonlinear systems have a quadratic-in-the-state representation. A sufficient condition is also shown in terms of differentially algebraic functions, which is useful to check the existence of a quadratic-in-the-state representation for most of practical systems.View full abstractDownload PDF (2112K) -
Toyonaka HIRAI, Yoshifumi KITAYAMA, Takao TAKAGI2003 Volume 39 Issue 12 Pages 1124-1128
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe past, the evaluation which affects human sensibility is vague, and, the fixed quantity was difficult. However, recently, the peculiar sensibility mechanism to these human is elucidated, and the research begins to advance as applied technologies of the retrieval, recognition, and the evaluation, etc. of the image.
In this report, We propose the method that five factors (change of color, color image, arrangement of pattern, change in shape, change in size) which exert the influence on the human psychology are gotten, and the technique by which nine factors (comfort, exhilaration, beauty, movement, showy, hot, weight, strength, hard) are presumed from those five factors. As a result, we could get a enough result to prove effectiveness.View full abstractDownload PDF (2735K) -
Seiichi OZAWA, Naoto SHIRAGA2003 Volume 39 Issue 12 Pages 1129-1135
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn reinforcement learning problems, the agent learns what to do so as to maximize numerical rewards. In many cases, the agent learns its proper actions through the estimation of an action-value function. When the agent's states are continuous, the action-value function cannot be represented by a lookup table in general. A solution for this problem is that a neural network is utilized for approximating it. However, when neural networks are trained incrementally, input-output relationships that are trained formerly tend to be collapsed by given new data. This phenomenon is called “interference”. Since the rewards are incrementally given from the environment, the interference could be also serious in reinforcement learning problems. To solve this problem, we propose a memory-based reinforcement learning model that is composed of Resource Allocating Network and memory. The distinctive feature of the proposed model is that it needs quite a small main memory to execute the accurate learning of action-value functions. To examine this feature, the proposed model is applied to the two conventional problems: Random Walk Task and Extended Mountain-Car Task. In these tasks, the learning domains are temporally expanded in order to evaluate the incremental learning ability. In the simulations, we verify that the proposed model can approximate proper action-value functions with quite a small main memory as compared with the conventional approaches.View full abstractDownload PDF (1816K) -
Takayuki HARA, Jun OTA, Hirofumi TAMURA, Toshimitsu HIGASHI2003 Volume 39 Issue 12 Pages 1136-1142
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, scheduling method is proposed for assignment of forklifts on an arrangement work. The proposed algorithm calculates a semi-optimal solution in practical time for the problem. This algorithm bases on simulated annealing and consists of three layers with a different search domain. We also propose online re-planning technique against the model error in offline plan stage, and the uncertainty of human's work. The validity of our method is verified in simulations.View full abstractDownload PDF (3219K) -
Katsuhiko FUWA, Tatsuo NARIKIYO, Hisao KAWAGUCHI2003 Volume 39 Issue 12 Pages 1143-1149
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn recent years, vibration control has been attracting great interest as an important research field on engineering. For example, in the case of control of the light weighted train, vibration caused by an influence on a disturbance and an appearance of resonances with flexible vehicle should be reduced for wide range of frequency. Furthermore, generally flexible structures have multiple resonance modes.
In this study, for a flexible beam with multiple vibration modes, we realize vibration suppression control system using disturbance observer to cope with frequency variation. In order to demonstrate the usefulness of the proposed disturbance observer, disturbance force whose frequency is varied for some range is applied to the flexible beam. It can be confirmed by numerical simulations and experiments that the proposed control scheme has superior performance of vibration suppression as compared with the traditional disturbance observer.View full abstractDownload PDF (1853K) -
Umerujan SAWUT, Gheyret TOHTI, Kazushi NAKANO2003 Volume 39 Issue 12 Pages 1150-1155
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSWith increasing demand for high fuel efficiency, better exhaust control measures, comfort and high performance on current vehicles, electronic control of the throttle chamber, which decides the engine output, is getting advanced day by day. This trend is expected to continue for future as well. Highly accurate and high performance systems can be constructed by the use of electronic control of the throttle chamber, in idle air quantity control, torque control, traction control, etc.
In this paper, precise modeling, identification, and position control of an electronically controlled throttle system, that has heavy nonlinear characteristics of gear, negative pressure, and spring, was shown. System identification was carried out using GA, and using disturbance VSS observers. Sliding mode nonlinear control was proposed for stable and better control of the identified system considering the practical application. The effectiveness of the control method was verified using simulation and experimental results on an experimental system.View full abstractDownload PDF (1447K) -
Wataru KASE2003 Volume 39 Issue 12 Pages 1156-1158
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSAn interactor matrix plays several important roles in the control system theory. Recently, we presented a simple method to derive a special interactor matrix using Moore-Penrose pseudoinverse. But, the structure of the proposed interactor was not specified. A triangular structure of interactor is useful for multivariable adaptive control. In this note, it will be shown a derivation of interactor with lower triangular structure. For this, a property of the interactor which we reported will play an important role.View full abstractDownload PDF (378K) -
Yasushi IWATANI, Masato ISHIKAWA, Shinji HARA2003 Volume 39 Issue 12 Pages 1159-1161
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, we propose a new kind of mobile robot called trident snake-like robot, which consists of a center body and three links having a free wheel. Its controllability structure is classified as ‘first-order’, i.e., the first level of Lie brackets together input vector fields span the tangent space. For the robot, we investigate a control method to steer the robot to an arbitrary position and posture.View full abstractDownload PDF (474K) -
Masayuki HARA, Takahiro HIGUCHI, Tetsuro YABUTA2003 Volume 39 Issue 12 Pages 1162-1164
Published: December 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSPerceptual illusion that is apparent in real-world environment does not result from human-error but from brain's functions that support perception. The present study demonstrated the effectiveness of our haptic interface to produce haptic illusion in virtual environment.View full abstractDownload PDF (1761K)
- |<
- <
- 1
- >
- >|