This paper presents some results obtained through analysis of a parameter identification problem for linear stochastic systems in the framework of functional analysis. A procedure is formulated for the identification of unknown parameters where a class of adaptation mechanisms is used to improve the filter dynamics so that the filtering residual provides an innovation process, i.e., a Brownian motion process independent of observation. The adaptation mechanisms are assumed to be modelled mathematically as a class of dynamical systems driven by the filtering residuals so as to provide a mapping from the function space made up of parameter estimation processes into itself. Sufficient conditions are obtained for the mapping to have a fixed point. The uniqueness of the fixed point is also proved to clarify the conditions for the parameter estimate, which is computed as the fixed point of a locally convergent adaptation mechanism, provides the unknown parameter asymptotically.

抄録全体を表示

This paper is concerned with the solvability and controller synthesis of the decentralized servo problem in the case where the outputs of each agent are required to track different types of reference commands without steady-state errors. Necessary and/or sufficient conditions for the solvability are derived. These lead to the fact that the conventional solvability condition for the robust decentralized servo problem is a sufficient condition for this problem. A synthesis procedure of decentralized servo systems is proposed based on these solvability conditions. It is also shown that the total order of the designed servo compensators may be smaller than that of the conventional one. Three numerical examples certify these results.

抄録全体を表示

This paper deals with the problem of designing an adaptive control system for discrete-time linear systems.
First, a design method of an adaptive control system for SISO systems where the relative degrees of the plant are unknown is proposed.
Next, the following three cases are considered.
(1) the plant has an unknown interactor matrix and can be decoupled by the state feedback control.
(2) the plant has partially a known interactor matrix and can't be decoupled by the state feedback control.
(3) the plant has an unknown interactor matrix and can't be decoupled by the state feedback control.
Simulation studies are shown to justify the proposed schemes.

抄録全体を表示

A feedback control scheme is considered for linear time-invariant systems with time delay. The control law used in this paper is the optimal control which is calculated for a quadratic performance index and the system obtained by removing delay elements from the given system. A sufficient condition for stability of the resulting closed-loop system is provided, which may be employed for computation of gain margin, gain reduction tolerance, and phase margin of the regulator. It is shown that when the closed-loop system satisfies the stability condition, the stability margins can be enlarged by increasing the feedback gain and thus a robust regulator for time-delay systems is constructed.

抄録全体を表示

In this paper, the position control system of a second-dimensional inverted pendulum is described. The second-dimensional inverted pendulum (s.d.i.p.) means the inverted pendulum with the fulcrum moving on the plane. The design of this control system consists of two problems. The first is the design of the driving mechanism of the cart which suports the fulcrum, and the structure of sensor to measure the angle of inclination between the vertical line and the pendulum. The second is controller's design. We use the Utmost Compensation Type controller as the controller to the s.d.i.p. In the last place, we present the experimental results.

抄録全体を表示

Real time 3-dimensional shape measurement is indispensable for online shape control system in steel industries.
No quick and accurate methods, however, have been proposed for time varying objects such as a running hot steel strip.
Moiré Topography is well known as a practical 3-dimensional shape measurement method, but its use has been limited to rest objects at room temperature.
Laser Moiré method, consisting of a Laser light source, a high-sensitive TV-camera and a heat-resisting grating, has been developed to apply Moiré Topography to online measurement in steel industries, and it is reported in this paper.
Using this method, high contrast and stable Moiré fringes were observed even for running hot steel strip at high temperature over 1000°C.
This method has made it possible to realize non-contact observation of the contour map of running hot steel over its total length.

抄録全体を表示

This report presents an investigation of the eutectic point of Ag-Cu alloy (Ag-71.9wt% and Cu 28.1wt%) as a reference temperature. The melting and freezing temperatures were measured by a platinum 10% rhodium vs platinum thermocouple and a high temperature platinum resistance thermometer. The melting temperature was determined to be t⁶⁸=779.46±0.02°C. On the other hand, the freezing temperature varied from 779.4 to 778.9°C, depending on the cooling rate. The mean value was t⁶⁸=779.13°C which was lower than the melting temperature by 0.33°C. Changing the concentrations of silver and copper by ±2.0% of silver from the exact eutectic point, significant differences were not evident for the melting temperature. In consequence of these results, it was concluded that the Ag-Cu eutectic point (melting) can be used as a practical reference temperature.

抄録全体を表示

Magnetic flowmeters should have non-conductive internal wall to prevent induced electromotive force from shunting. Most of commercial flowmeters are lined by insulating material. The lining material limits applicable temperature range of measured fluid and also reliability of meters. We have alrealdy shown that the lining can be substituted by the wall potential formed by applying electric current externally. The principle of the proposed magnetic flowmeter is to keep the potential distribution over the inner wall almost identical to the flow induced potential in the liquid so that no current flows across the boundary between fluid and wall. Then the output signal becomes equal to that of conventinal magnetic flowmeters.
In this paper, the proposed magnetic flowmeter is theoretically analyzed. A potential problem for wall (U_w) and fluid (U_f) is formulated as; ΔU_f=div(v×B), ΔU_v=0 subject to continuity equations for potential and current at the boundary between metalic pipe and fluid and that a current distribution normal to outer wall surface is equal to the applied current distribution. This problem is solved explicitlly under the assumptions: (1) point signal pickup electrodes, (2) linear current feeding electrodes, (3) an axially symmetric flow profile and (4) much higher conductivity of metalic pipes than fluids.
Theoretical solutions agree well with experimental results already presented. They also give the optimal configuration of current feeding electrodes and the optimal value of the control parameter of electrical servo-system for potential formation. When current electrodes are placed at the angle 64.5 degrees and other symmetric positions and the control parameter is set at 0.981, the output signal depends fluid conductivity not at all. Therefore the formed potential distribution compensates wall conductivity effect perfectlly.

抄録全体を表示

For the purpose of developing a compact and light weighted measurement system for the concentration of electrolytic solution, a new technique with the application of transmittancy membrane is devised.
It is known that a liquid densitometer (conductivity meter) is commonly used for the measurement of electrolytic solution concentration. For the purpose, there are DC and AC (below several KHz) bridges and either of which has merits and demerits. Especially when AC bridge is used, apparent solution impedance obtained is containing both surface and solution impedance. In order to eliminate the influence of the surface impedance, it is better to increase the measurement frequency up tot some MHz order, however, RF bridge is not easy to operate and is apt to include some stray error. A resonance method by a Q meter for measuring RF impedance is the most easy way.
This is the way of utilizing characteristics of a solution transmitted and diffused into pure water through a control plate and cellophane membrane since the solution concentration varies in the pure water. Namely, continuous measurement by a Q meter has been tested regarding Q variation of a detector coil which is installed in pure water. The most remarkable feature of this method is an indirect measurement by diffusing and transmitting solution into pure water through membrane rather than direct measurement on the solution. It is possible to obtain many kinds of information. A spiral coil which is printed on a board by photoetching technique is used as a detector. It has several advantages that the sensor becomes rather small and light weighted so that it occupies less space which turned out less quantity of solution under sample and the coil parameters are stable.
Experiments have been conducted for various types of solution and for the extent of influence of the control plate. It has been found out that each solution has a proper curve for different value of concentration with the optimum condition of a control plate. Analysis on each curve resulted fundermental characteristic values which can be used for the identification of solution.

抄録全体を表示

A microcomputer based measurement system of 3-D motion has been devised using a pair of position sensing detector (PSD) cameras, which monitors movements of LEDs attached to various places on a moving object. The system employs a novel method for determinations of camera parameters (position and orientation of each camera relative to the objective space) by displaying an LED moving along three lines of known orientations in the space in advance to measurements, which effects to exclude measurement errors due to inaccurate presetting of each camera at a specified position. It also enables to set cameras on any appropriate places so that they can monitor whole movements of LEDs without being hidden by other part of moving object. Resultant determination error of orientation of camera axis proved less than one third of a degree.
Other various techniques are employed to improve measurement accuracy and to reduce noise, including real time correction of spatial distortion of position outputs of the PSD cameras. Average measurement error and minimum detectable displacement (noise) proved less than 0.5mm over central cubic region (with 40cm edges) and less than 1.0mm over peripheral region (inside a cube with 60cm edges excluding the central region) with an inter axes angle of 60° between cameras and with a distance of 2m between an origine of the space and each camera. Time for measuring a point and calculating its position are 3.5msec and 230msec, respectively.

抄録全体を表示

One of the major problems in designing the magnetic bearing control system is that the two rotational motions of the rotor interact on each other by gyroscopic effects. Two control methods are presented for the gyroscopic coupled system. The system compensated by one method, which is called optimal regulator system, has an antisymmetric cross feedback compensator. The system compensated by the other method, which is called independently controlled system, has no cross feedback compensator. There are many differences in dynamics between the two compensated systems. The differences are explained clearly by representing the system dynamics in complex variable form. And it is shown that the optimal regulator system is superior to the independently controlled system in point of transient response characteristics.

抄録全体を表示

This paper deals with tank tests for models of ferryboat to cross a river without any fuel. The boats are expected to use in rivers of Himalayan areas, and their sailing power is given by running water of the river.
The principle of the ferryboat is as follows; A rope longer than width of the river is used for the ferryboat. The rope is connected between the bank and the boat, and inclines the boat against the flow direction. Then thes boat will sail to the opposite side of the river in a radius of the rope length, the principle is similar to a kite in the sky.
The four models are used for the tank test. A typical boat is a rectangular prism with a base 0.1m×0.22m and a hight of 0.04m. The other three boats cut the bow 30, 45 and 60 degrees respectively against horizontal plane.
Since each boat has a flat bottom, aluminum pressure plate with area 0.1m×0.03m and thickness of 0.001m is attached to the bottom, so as to increase the capacity of crossing the river.
As the results of the tank tests, it made clear that;
1) The rectangular prism which has the biggest drag went down the front half part of the boat. This is due to the drag.
2) The boat having the smallest drag with the pressure plate obtained the angle of 57 degrees between the bank and the rope. Which means that the rope length requires only 1.3 times of the river width. This has been confirmed at the Tone, Arakawa and Tenryu rivers by means of the real boats.

抄録全体を表示