We had proposed and developed an electromagnetic thickness sensor which can measure the thickness of a conductor plate or tube coated with rust. However, our measurements were made under rust-free conditions in the previous paper. Therefore, this paper presents the measurement of simulated rusty objects. The theoretical analysis related to rust-free measurement is expanded accordingly. The effects of rust on measurement are discussed using this expanded analytical result. Experimental results for steel plates coated with Teflon film in place of rust verify that our method can measure the thickness of rusty objects and confirm that the upper limit of the measurement is about 2.5mm when the Teflon film thickness is 0.5mm.

抄録全体を表示

This paper proposes a new method to estimate time delay of a random impulse sequence in a short time with high accuracy.
In the proposed method, a Reset-Set Flip-Flop is used to obtain the rectangular pulse sequences which are set by the impulses of original signal and reset by the ones of delayed signal. Then the time delay which is sufficiently smal than the mean interval between impulses of the signal is directly estimated by measuring the total width of rectangular pulses in the given short time.
However, if the time delay is not so small then the output rectangular pulses of the Flip-Flop are incorrectly reset by the impulses coming before the real delayed ones, hence the estimated time delay becomes smaller than the real one. In order to overcome this disadvantage we propose the employment of adaptive gate which prohibits the inpouring of the impulses coming before the real delayed ones into the Flip-Flop. By this means, the output rectangular pulses can be reset correctly and we can estimate the delay accurately up to any desired large range.
The validity of the presented processes are verified by computer simulations using the Poisson random impulse sequences.
This method is also applicable for continuous random signals if we transform them into the corresponding pulse sequences by using proper pre-processing. The results applied to the continuous signals which have been obtained from the actual instruments showed the usefulness of the proposed method. The result was as good as that of the second-order cross-correlation method.

抄録全体を表示

A simple method is presented on how to generate the state feedback which makes the closed loop transfer matrix the inverse of the interactor matrix. This can be regarded as a generalization of decoupling control using state feedback. An arbitrary pole assignment which preserves the lower left triangular form of the inverse of the interactor matrix is also considered.

抄録全体を表示

In this paper, it is shown that Kalman's conjecture is valid for discrete nonlinear feedback system if Hurwitz sector is replaced by more restricted sector that is called ESF-gain sector. Many of practical nonlinearities are included in this class. It is also shown that ESF-gain sector which depends on the solution of Lyapunov's matrix equation can be derived by solving a quadratic equation.

抄録全体を表示

The paper considers two methods for active control of large flexible structures using pole assignment for the purpose of stabilizing the neglected dynamics of the unmodeled higher modes of vibration. The first one based on pole placement in a vertical strip has the merit of minimizing control forces and infusing damping in a wide range of modal frequencies. The second approach based on using linear state feedback regulator with Kalman filter observer also uses pole placement in a prescribed region through a quite natural and easy design of the regulator and observer. Both methods have been applied for active stabilization of a flexible feed-support structure for a large parabolic antenna.

抄録全体を表示

This paper is concerned with the computation of Nash solutions for a class of deterministic linear-quadratic two-person nonzerosum difference games. The dynamics of the system is characterized by a linear discretetime descriptor equation Ex_k+1=Ax_k+Bu_k+Cv_k, where E is singular. The games are studied under the open-loop information structure. It is proven that the linear quadratic Nash game of the discrete descriptor system always admit uncountable Nash equilibrium solutions even if the information structure of the game is open-loop. This property is very different from that in the open-loop Nash game of a state space system. A pair of minimum-norm open-loop Nash solutions is computed by recursively solving two coupled asymmetric Riccati-type matrix equations.

抄録全体を表示

This paper introduces a control method for biped walking on rugged terrain. We assume a biped robot which has a mass body and massless legs, and which is restricted to move in a vertical plane. We apply constrain control to the body of the robot so that it moves on a particular straight line and rotates at a constant angular velocity. As the result of such control, the dynamics of the center of mass of the body becomes completely linear. We also notice that the dynamics of the body does not depend on the slope of the line on which the center of mass of the body is constrained. We call such motion of the ideal model “Linear Inverted Pendulum Mode”. The limitation of the condition under which this method works is also discussed.
After these consideration, we develop the control scheme of the biped walking on rugged terrain. Under the control based on the linear inverted pendulum mode, walking on a particular rugged terrain is shown to be equivalent to walking on a level surface.
Finally, it is shown that by using ankle torque, our method is robust and can be applied to a real biped robot with mass legs. That is ascertained by simulation.

抄録全体を表示

In a compliane control of a robot manipulator, an independent joint control approach is quite important because its features include: (1) simple SISO controllers, (2) collocated sensors and actuators. In a previous paper, we have proposed a method named “Stable Negative Compliance Control (SNCC)” based on such an independent control approach, whereby arbitrary end-point compliance can be achieved in arbitrary arm configurations.
In this paper, stability analysis for the SNCC are first discussed. When the SNCC is realized by an ideal torque feedback and nonlinear dynamic terms of the arm can be neglected, the control system remains always stable in a free space. Comparing conventional Direct Compliance Control (DCC) and Salisbury's Active Stiffness Control (ASC), it is shown that SNCC only can realize arbitrary symmetric end-point compliance matrix. Next, the control schemes for the Stable Negative Compliance (SNC), which is based on the characteristics of an inverse response of non-minimum phase shift system, have been experimentally confirmed by a 1 DOF actuator system. Although the controller has one unstable pole, the overall closed loop can be easily stabilized and provide an arbitrary SNC value. Finally, the SNCC is implemented on a real planar 3-link direct drive arm. Featuring the decoupled end-point compliance, the validity of the method is demonstrated by a real task.

抄録全体を表示

An artificial satellite which has a magnetically suspended momentum wheel is expected to have high reliability and accurate attitude-control-performance. This type of wheel can also be used as an actuator in directions transverse to its spin axis, by controlling the tilt angle of the wheel rotor actively. So the three-axis control of a satellite is possible with a single wheel.
However, the small stiffness of magnetic bearings can cause the instability of the satellite-wheel system and make the nutation of the satellite grow.
In this paper, a new feedback control method is proposed to stabilize the nutation of a satellite. That is, the nutation of the satellite is detected by measuring the control torque of the magnetic bearing, and the control torque is added to the gimbal command through a low pass filter.
The effectiveness of the proposed method is verified analytically by using Routh row and confirmed by the computer simulation.

抄録全体を表示

In the combined cycle power generation plant for coal gasification, it is necessary to establish a coordinated control for the gasifier and the gas turbine according to a load. Therefore, a dynamic model which expresses system's dynamic behavior with regard to various kinds of coal and operating conditions is required.
In this paper, a dynamic model dealing with an air blow pressurized two-stage entrained bed gasifier is first considered. And then, it has been confirmed that this model is valid by comparing the dynamic responses of this model with those of 2 tons/day bench scale gasifier.
In the entrained bed gasifier, intricate chemical reactions predominate between coal particles and gases in flowing state and hence it is difficult to describe the reactions exactly. Since our object is to control the gasifier's operation, the model has been so simplified as to fully indicate the behaviors of the air blow two-stage entrained bed gasifier (the gasifier has been lumped into two parts- upper and lower parts-, and have taken only the reaction of carbon, hydrogen and oxygen have been take into acount).
The results of the dynamic simulation of the present model are as follows: in the entrained bed gasifier, the residence time of coal particles is very short (about ten seconds), and the responses of the produced gas property, the gasifier temperature and the gasifier pressure are extremely fast corresponding to the change of coal-feed rate or air-feed rate. From these results, it is important to control the coalfeed rate, the air-feed rate and the gasifier pressure for the stable and sustained operation of the gasifier.

抄録全体を表示

A fuzzy control has a distinguished feature that the control is capable of incorporating expert's control rules using their linguistic descriptions. Several researches have been done on the learning capability of neural networks for their applications of fuzzy control. But, there has been no controller which can automatically identify the rules in the consequences and tune the membership functions in the premises.
This paper presents a fuzzy controller using a neural network. The new controller can automatically identify the control rules and tune the membership functions utilizing expert's control data. Identification capability of the new fuzzy controller is examined using simple numerical data. The results show that the network can identify nonlinear systems to the same extent of precision as conventional fuzzy modeling methods do. And for demonstrating the capability of the new fuzzy controller, we simulate step responses of a control system which incorporates the fuzzy controller and a controlled object of a first-order system with a dead time.

抄録全体を表示

Many local cities are groping for plans to activate themselves. One effective way is to carry out large-scale projects in the cities, and then by impact of the projects is to achieve the change of their economic structure and the extension of their economic scale. Accordingly, Some tools for quantitatively assessing effectiveness of the projects are needed to make successful plans.
In this paper, we describe an impact metric model of large-scale projects to meet such needs. The model consists of a regional econometric model, a regional input-output model, and population model. The feature of the model is to be able to measure not only impact of direct investment of the projects but also one of working the projects. Moreover, a simple method is proposed to mesure impact of consumption cycle (production-income-consumption-production) of the projects. Finally, we apply the model to Kitakyushu city and then present some simulation results.

抄録全体を表示

In this paper, we review the essence of the “double measuring method” and propose the “improved double measuring method”. Some advantages of the improved method have been pointed out.

抄録全体を表示

An analysis and evaluation of the digitization error of area measurements in digital image mensuration are presented. It is shown that so called round-off error in digitization process of figures can be evaluated in proportion to their perimeters and N^-3/2, where N is the digitization lattice density. Theoretical and experimental results of digitization error analysis for circular figures are presented.

抄録全体を表示

This paper describes the acoustic inspection for working states of burnishing brush with the neural network. Here the two cases have been considered as the input data for neural network. One is the acoustic signals from brush and the other is their spectrum data.

抄録全体を表示

A recent paper has presented a discrete-time repetitive control scheme for periodically time-varying linear systems. The resulting closed-loop system has characteristic poles specified by a designer. However, it has been assumed that the plant is stably invertible. This paper proposes a repetitive control algorithm applicable to unstably invertible periodic systems.

抄録全体を表示

We discussed the problem of designing an observer for linear multivariable descriptor systems with unknown inputs. New sufficient conditions are found, under which the observer should exist, and a simple designing procedure is presented.

抄録全体を表示

In this paper we derive inverse Nyquist criterion for systems with time-delay. It can be applied when nominal model and perturbed model have same number of zeros, instead of poles, in right half plane.

抄録全体を表示

This paper deals with an attitude control system synthesis for a large satellite. Two step design procedure is proposed which consists of the velocity and position loop. This method is considered as a design method for the DVFB (Direct velocity feedback) system. The effectiveness of the method is examined through control experiments.

抄録全体を表示

This paper considers a PBH test for reachability in a linear time-varying discrete-time systems. Let R denote the set of all real-valued functions defined on the set of integers. We consider a time-varying discrete-time system to be over R, and characterize the system in terms of a semilinear transformation. Mode-reachability is defined in terms of eigenvalues and eigenvectors of the semi-linear transformation. It is shown that mode-reachability is equivalent to Kalman's original reachability.

抄録全体を表示

The design method of the feedforward rapid position control system of a beam structure based on the finite element method and inverse dynamics is presented. From the experimental results, it is confirmed that the beam structure can be controlled rapidly.

抄録全体を表示