-
Tohru IUCHI, Tomoyuki TSURUKAWAYA, Akira TAZOE
1998 Volume 34 Issue 3 Pages
175-181
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
Emissivity of a metal is fairly low in general. Due to an oxidation film formed on the metal surface, however, its emissivity changes rapidly, and large errors in the temperature measurement can be caused. This is a serious problem for the radiation thermometry of metals. In order to eliminate the error due to surface conditions of the targets, the authors propose a new method based on their findings that the apparent emissivity of a metal surface depends on the direction from which the target is looked at. This means that the radiance of the target does not obey Lambert law. The proposed method uses the ratio of radiance measurements taken at two different angles to compensate the emissivity or the emissivity ratio. This method is promising for the in-line use at metal manufacturing processes such as steels or aluminums.
View full abstract
-
Ippei TORIGOE, Yasushi ISHII
1998 Volume 34 Issue 3 Pages
182-187
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
A principle is proposed for measuring the surface area of an object. The heat flow between an object and the surrounding medium is proportional to the surface area of the object. On this heat conduction law the surface area of an object can be measured. In this paper one of the possible embodiments of this principle -a method utilizing the acoustic impedance- is described. For sufficiently low frequencies, the acoustic impedance of a cavity of an adiabatic wall is capacitive. For real cavities, however, a real component arises from heat conduction at the wall. And the magnitude of this component is proportional to the area of the cavity wall. Therefore the surface area of the cavity, or of an object placed in a known container, can be known from the acoustic impedance measurement. In the first place, the acoustic impedance of a cavity is calculated assuming idealized conditions and the relationship is obtained between the cavity surface area and the acoustic impedance. In the second place, the influence to the acoustic impedance of the departures from the idealized conditions are investigated: the effects of the wall properties and the shape of the cavity. Finally, other influence quantities are enumerated and discussed summarily.
View full abstract
-
Masafumi OKADA, Toshiharu SUGIE
1998 Volume 34 Issue 3 Pages
188-194
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
Recently, the importance of the joint design of identification and control has been recognized, and several controller design methods based on the iteration of identification and controller re-design have been proposed. In these methods, the frequency weighted identification plays an important role. On the other hand, as a powerful identification method, subspace state-space system identification (4SID) method has been proposed. However, it is difficult to use the frequency weight in conventional 4SID methods. Therefore in this paper, we propose a frequency weighted 4SID method for the joint design of identification and control, and choose the appropriate frequency weighting function for identification considering the cost function of the controller design. Furthermore, the effectiveness of the proposed method is evaluated by numerical examples.
View full abstract
-
Yuuki KUROE, Takao FUJII
1998 Volume 34 Issue 3 Pages
195-202
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
ILQ design theory for optimal servo systems based on the result of the inverse LQ problem has been generalized so far with respect to both reference inputs and controlled objects. In this paper we integrate several existing results of ILQ design theory to propose a unified design method of optimal servo systems which can treat nonminimum phase systems with direct transmission as well as generalized reference inputs. First, we extend the usual decoupling method and the two characteristic structures of ILQ optimal servo system to such systems. Next we use them to obtain a practically useful design method of two ILQ feedback gains: one of the gains called “ILQ nominal feedback gain” is expressed analytically in terms of system parameters and design specifications. The other gain called “ILQ tuning parameter” is determined so as to guarantee LQ optimality of the servo system. Finally we show a numerical example to illustrate the validity of the method.
View full abstract
-
Shape Control Based on Curve Parameter Estimation
Hiromi MOCHIYAMA, Etsujiro SHIMEMURA, Hisato KOBAYASHI
1998 Volume 34 Issue 3 Pages
203-208
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
In this paper, we propose a new shape control law based on curve parameter estimation for a given parametric curve. The proposed control law includes a very simple and geometrically meaningful estimation law which infers curve parameters corresponding to target joint positions on a curve. Due to this control law, we are able to control shapes of a manipulator without solving any complex nonlinear equations numerically to obtain target curve parameters before starting the control, which is needed for conventional PD feedback control in task space.
View full abstract
-
Yoshihito IMAI, Teiji TAKAHASHI, Atsushi MORITA
1998 Volume 34 Issue 3 Pages
209-216
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
Feedback control has been used to stabilize the steady-state output power of a CO
2 laser to overcome the problems caused by the change in the temperature/deterioration of CO
2 gas. The transient response, however, is as slow as a few hundred milliseconds because of the slow dynamics of a thermopile power sensor. When machining conditions of a CO
2 laser are changed, this rather slow response requires an extra dwell time, resulting in low productivity of the machining.
To cope with this problem, the authors have developed adaptive feedforward control for a CO
2 laser in addition to conventional feedback control. The model of a CO
2 laser is described as a gain, which is varied by the setting parameters; laser power, pulse frequency and duty factor, as well as gas conditions. In this paper, two new variables, effective discharge power and threshold discharge power, are introduced to obtain a compact and adjustable model. With the proposed control system, the step response time of a laser power is reduced to less than ten milliseconds without overshoot, and can be set to desired constant time. The effects of such a fast and stable response on the machining speed and machining quality are examined. The experimental results show that for thin metal line-cutting, neither the melt-off area nor dross is observed even in the no-dwell time case. For thin metal hole-cutting, the machining speed is improved by 30%.
View full abstract
-
Kazuo YOSHIDA, Shinji SETO, Toshiaki MAKINO
1998 Volume 34 Issue 3 Pages
217-224
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
It becomes difficult for a current collector of high-speed railways to maintain constant contact force between the contact strip and the trolley-wire, since it suffers strongly from wire and lift disturbances in high-speed. Accordingly, it is necessary to control the height of current collector. Electrical sensors can be installed only below the insulator, since the upper part of the insulator is located in high voltage. In this paper, it is clarified to be difficult to distinguish wire disturbance from lift one in low frequency range. However, this difficulty is solved by separating the frequency range of these disturbances. Two kinds of contact force control using
H∞ control theory are presented in this paper. One is a servo
H∞ controller of contact force using the lift estimated from drag. This controller can reduce the contact force variation due to wire disturbance at the frequency range less than 1.5Hz, and the contact force due to the lift can be reduced if the lift is estimated exactly through the drag force. The other is a
H∞ regulator which estimates those disturbances in the separated frequency ranges. This controller can reduce the contact force variation due to the wire disturbance at the frequency range from 0.2Hz to 6Hz compared and contact force variation due to the lift at less than 0.1Hz. Furthermore, it is confirmed that if a sensor can measure the displacement of contact strip, contact force can be controlled better against those disturbances.
View full abstract
-
Shigeto OUCHI, Kang Zhi LIU, Shouji SATO, Tsutomu MITA
1998 Volume 34 Issue 3 Pages
225-231
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
Mine train control system is consisted of a winder, a train which transports many people to minebottom, and a long wire rope which connects a winder and a train. The mine train control system reported in this paper is the system with the longest rope of about 6700m in the world. This system often generate a rolling of train when the train stops at the station, when using conventional control. In convensional systems, two mine train control systems of the length of 1/2 of all mining galleries were used in order to ristrict the rolling. In these systems, change time was necessary. However, this change time become unnecessary by using a system with the length of 6700m. When using this system, it is the main subject how to control the bigger rolling. We solved the problem on the rolling control of this system by using
H∞ control, and got the result that the travel time of the train made 2/3 compared with that of the convensional systems using two mine train control systems, and the train speed of 420m/min which is the maximum speed in the world was achieved.
View full abstract
-
Koichi HIROTA, Toyohisa KANEKO
1998 Volume 34 Issue 3 Pages
232-238
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
In the practical application of virtual reality technology, such as surgical simulation, feedback of force is strongly needed and various force feedback devices have been developed. However, for the interaction with the virtual object using force, not only the interface device but also the kinetic model of the object must be implemented. In this paper, calculation methods of elastic virtual object model are discussed.
The simulation algorithm of the transformation can be divided in two parts: the transformation model and the solving method. In our algorithm, we combined the physical elastic model with the differential solving method. The physical elastic model allows us to define the elasticity of the model by physical parameters, and the differential solving method provides us short simulation cycle time that is essential to the real-time interaction interface.
A prototype environment to interact with the proposed model through 3d pointing device was developed. By the prototype, real-time operations such as pulling and cutting on to the model was proved to be possible.
View full abstract
-
Yoshio MOGAMI, Norio BABA, Masaki MATSUSHITA
1998 Volume 34 Issue 3 Pages
239-246
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
In this paper, a new learning algorithm for variable hierarchical structure learning automata operating in a P-model stationary random environment is constructed by extending the estimator algorithm proposed by Thathachar and Sastry. The learning propertiy of our algorithm is considered theoretically, and it is proved that the optimal path probability can be approached 1 as much as possible by using our algorithm. In numerical simulation, the average number of iterations of our algorithm is compared with that of the variable hierarchical structure learning algorithm of
LR-I type proposed by Mogami and Baba, and it is shown that our algorithm can find the optimal path after the smaller number of iterations than that of the algorithm of
LR-I type.
View full abstract
-
Takeaki AKATSUKA, Nobuo FURUMATSU, Hirokazu NISHITANI
1998 Volume 34 Issue 3 Pages
247-253
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
When we design a new plant or plan a production schedule, simulation is the most practical technique to forecast and evaluate the plant operation. The chemical processes in a plant with control systems are represented by a set of differential equations and algebraic equations. Continuous simulation can solve these equations. However, the plant is usually controlled by the sequential control system, and is sometimes operated by the human operators in the discrete event mode. These facts require a combined discrete-event and continuous-process simulation. In this paper, we developed a new simulation method which more properly fulfills the functions necessary to simulate the combined discrete-events and continuous-processes. In the method, the target process is modeled by three types of elements, i.e., continuous-process, discrete-event, and sequential control. Since sequential control has the intermediate characteristics between discrete-event and continuous-process, it is more convenient for modeling to handle it as the third element. The simulation program can provide a powerful tool to simulate the complex systems such as a combined batch and continuous process, a pipeless batch plant, a textile mill and so on.
View full abstract
-
Parallelization of Expressions by Algebraic Transformation
Kiyoharu TAGAWA, Takahiro NAMIKOSHI, Hiromasa HANEDA
1998 Volume 34 Issue 3 Pages
254-262
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
This paper presents a new task-generation method for robot-arm control computation on a multi-processor system with an arbitrary number of processors. In order to execute the effective parallel processing which achieves high speed-up ratio, the definition of optimal task-set is an essential problem. The proposed task-generation method, which is based on algebraic computation techniques, consists of three phases: 1) Derive a robot-control law by using symbolic languages, such as Mathematica or REDUCE, as a set of expressions; 2) Transform the expressions into a simplest possible form which does not contain any redundant operations. In this phase, the factorization algorithm which was recently proposed by the authors is used; 3) Divide each equation into subexpressions repeatedly to induce the parallelism of computation. Consequently, each of the optimized expressions is regarded as one task. By using the proposed task-generation method, an automatic parallelizing compiler is also developed. Experiment on the inverse dynamics computation of a six-joint robot-arm demonstrates the effectiveness and the usefulness of the proposed task-generation method.
View full abstract
-
Atsushi AOKI, Hitoshi IIMA, Nobuo SANNOMIYA, Yasunori KOBAYASHI
1998 Volume 34 Issue 3 Pages
263-269
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
This paper deals with a job shop scheduling problem with a multi-function machine. In this problem, the single function machine processes the specific operation of each job, and the multi-function machine can process all operations of any job. Therefore each operation must be determined so as to be processed by either the single function machine or the multi-function machine. The objective function of the problem is to minimize the completion time. A genetic algorithm is applied for solving this problem. Two procedures are proposed to revise the algorithm for dealing with this problem. One is to define a new chromosome in addition to the chromosome used in the conventional job shop problem. The other is to use a heuristic method for decoding the chromosome. The effectiveness is compared between the procedures by examining numerical results. Consequently, using a heuristic method is more effective than using two chromosomes.
View full abstract
-
Yoshifumi KITAYAMA, Takao TAKAGI, Yuzo NAKASHIMA
1998 Volume 34 Issue 3 Pages
270-274
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
The impression which an illustrations in newspaper advertisements and wrapping paper give to human beings plays an important role in commodity sales. When we try to evaluate the comfort level which these illustrations give to human beings, 1/f fluctuation caused by the colour array can be considered as one important element. As a fundamental study of the method to evaluate such illustrations from the standpoint of a colour array, we have studied a method that measures the comfort level of round shaped four colours with conformity arranged on plain white paper.
This system presupposes that the four colours exert a mutual influence of potential energy according to their arrangement. We have tried to define and evaluate this energy. The comfort level is measured using the calculation value of energy. For this study, we have used the above proposed system to compare 16 kinds of samples conforming to the above method calculating the value of energy with the results of a questionnaire answered by two or more groups of human beings.
View full abstract
-
Yoji YAMADA, Masaru NISHIGAKI, Sheng Yang HUANG, Yoji UMETANI
1998 Volume 34 Issue 3 Pages
275-277
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
This short paper concerns development of dynamic tactile sensor elements each of which can be independently distributed on a compliant skin. Through-hole downward wiring structure of the head of the sensor elements gives no mechanical interaction among them and makes them possible to be arranged even on curved surfaces. 3-axis force sensing and slip sensing functions of the sensor elements are achieved basically through the contact structure beteewn a quadrangular pyramid tip and a base shaped engraving, which is advantageous in making the size of the sensor elements very small. We utilized PVDF film transducers for the above sensing capabilities, fabricated sensor elements with the proposed structure, and evaluated the performance of a sensor element for verifying its sensing principles.
View full abstract
-
Osamu MITSUI, Hitoshi IIMA, Nobuo SANNOMIYA
1998 Volume 34 Issue 3 Pages
278-280
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
This paper deals with a scheduling problem in a flow shop process with parallel machines. The objective of this problem is to minimize the makespan, and the problem has complicated constraints. We propose two kinds of genetic algorithm to solve the problem. By examining numerical results, a comparison is made between them from the viewpoint of accuracy.
View full abstract
-
Tadanari TANIGUCHI, Kazuo TANAKA
1998 Volume 34 Issue 3 Pages
281-283
Published: March 31, 1998
Released on J-STAGE: March 27, 2009
JOURNAL
FREE ACCESS
A fuzzy descriptor system is defined in this paper. Three kinds of stability conditions for the fuzzy descriptor system are derived and represented in terms of linear matrix inequalities (LMIs). The stability analysis is reduced to a problem of finding a common Lyapunov function. An LMI design approach is employed to find stable feedback gains and a common Lyapunov function.
View full abstract