A slip sensor using a conventional pressure gauge is developed to realize a robot hand which is capable to hold soft objects, such as a paper cup, pudding, etc. The slip sensor consists of an acrylic block with a cylindrical hole, of which one end is shielded with a rubber sheet and a pressure gauge is mounted on the other end. Water is filled in the hole with a little higher pressure than the atmospheric to swell out the rubber membrane. The movement of an object which contacts with the membrane makes stick-slip vibration and in consequence it makes an impulsive pressure change of the water, superimposed on the static pressure. By observing the impulsive pressure changes detected by a pressure gauge, a slip of a smooth acrylic surface by the velocity of 2.5mm/s against the membrane could be detected. A robot hand using this slip sensor could hold a paper cup with time varying water content, soya bean pudding, etc.

抄録全体を表示

A new method based on an acoustic approach is described for locating a pinhole occurred in a gas pipeline. Increase or decrease in the pressure in a packed pipeline with a pinhole leads to the gas blow out or in from the pinhole by which sound noise with the wide frequency range is generated. The acoustic noise generated at the pinhole induces the standing waves in the pipeline just like as a flute. Modes of the standing waves are determined by the length of the packed pipeline and the site of the pinhole.
A mathematical acoustic model to describe the phenomena above was built and a method to locate a pinhole was presented by using the model. The method requires only one microphone which is set at a terminal in the packed pipeline. A simple algorithm applying the Fourier transformation to the acoustic signal detected by the microphone yields the accurate location of the pinhole site of the pipeline. Laboratory experiments verified the modeling and the location method.

抄録全体を表示

As a functional type fiber temperature sensor has a long sensing fiber for a range of measurands, the phase sensitivity is high for temperature change. An optical heterodyne interferometer has been used to measure small phase variation in the temperature sensor. However, the interferometer has suffered from external disturbance because the reference light has a different path from the measurement arm. This paper describes an optical fiber temperature sensor using heterodyne interferometory to enable free from external disturbance easily. The measurement arm is made loop to remove the reference arm. A stabilizied transverse Zeeman laser (STZL) is used to exclude the phase fluctuation due to the phase difference between the measurement path and the reference path. The STZL has the leak frequncy components. The leakage causes the periodic nonlinearity of the heterodyne interferometer and decreases the accuracy of the interferometer. The relation between the cross talk due to the leakage and the measured phase error is analyzed with considering initial phase difference between polarization components of the STZL beam. From the analyzed result it is found that the cross talk can be measured from amplitude modulation of beat signal. To reduce crosstalk term, the leakage is decreased by phase compensation method using a quarter-wavelength plate, and the cross talk term is cancelled by a new fringe distribution method. In the temperature sensor the reduction of the non-linearity is confirmed experimentaly.

抄録全体を表示

A sensor by means of optical fiber bundle was developed for detecting microscopic displacement and vibration in tympanic membrance. Illuminating a planer object and detecting the reflected light with the bundle, we experimentally investigated the received optical intensity against the distance from the object. As a practical application to clinical examinations, we then measure the displacements and vibrations of tympanic membranes of a dog.
The main results of experiments are as follows.
1) The reflected intensity increases rapidly with the distance in the front slope region at a small distance, then it reaches the maximum at a certain distance. As the distance increases further into the back slope region, it decreases linearly. Straight line portions of such a intensity-distance characteristics enable us to obtain inversely the displacements and vibrations from the intensity measured.
2) The displacement resolution is 0.05μm. The sensitivity is 9.85 mV/μm for the region with L=0.1-0.5mm (front slope region) and 7.97 mV/μm for L=1.0-1.7mm (back slope region).
3) The frequency response of dog tympanic membrance was successfully measured for frequency range with 100-8, 000Hz. We conclude that the fiber optics bundle well detects the microscopic displacements and vibrations in tympanic membranes and its application to clinical examinations are expected.

抄録全体を表示

A robust servosystem design using the method of inequalities is proposed for single-input single-output plants with poles in the open left half plane and/or at the origin. This type of plant is ubiquitous in various control fields. We obtain a simple pole-zero placement algorithm which has some relation to Q-parametrization of all the stabilizing controllers for asymptotically stable plants. This algorithm is used to construct a CAD system based on the method of inequalities. Unlike CAD systems using YJB-parametrization, this system can be used without a tentative stabilizing controller even if a plant is not asymptotically stable. We propose a reasonable choice of the tuning parameters in the controller and efficient numerical methods for the performance evaluation. A prototype of the CAD system is constructed on a personal computer. Design examples are given to illustrate the effectiveness of the CAD system.

抄録全体を表示

A ripple-free deadbeat control system is one in which the controlled output is driven to a step reference input and eliminates effect of a step disturbance in finite time with no ripple. Deadbeat control is said to have a defect in transient response.
This paper presents a synthesis of ripplefree deadbeat control system with optimal transient response. The synthesis is simple and plain, for the optimization problem formulated in this paper has no constraint and an explicit solution is obtained.

抄録全体を表示

Unstable zeros in discrete-time models and computation delays are key issues in digital controller design. To cope with these issues, we consider an application of the partial loop transfer recovery technique which has been proposed by Moore and Xia for continuoustime non-minimum phase systems. Taking account of computation delays, we construct a discrete-time LQG controller that depends only on the estimates of the minimum phase states of a plant model. The controller consists of a dynamical feedback gain matrix, a minimum phase state predictor and a Kalman filter estimating the minimum phase states. We consider feedback properties at the input of the plant. It is shown that, by increasing the variance of a fictitious disturbance injected into the minimum phase part of the plant model, we can recover feedback properties achieved by the minimum phase state feedback. To prove the recovery, we use simple decompositions of the sensitivity matrices. We give an explicit representation of the recovered sensitivity matrix. A numerical example is presented to illustrate the effectiveness of the proposed design.

抄録全体を表示

This paper proposes a model reference adaptive control system (MRACS) for any SISO, continuous-time, minimum-phase linear plants. From the thoretical viewpoint, the only required a priori information about the plant is that the upper bound of the plant degree be known. The exact value of the plant relative degree is not required.
Since the upper bound of relative degree of the plant is finite and is known, we can consider the finite number of parameter adjustment rules, each of which is constructed from the possible relative degree and the possible sign of high frequency gain of the plant. The adaptive algorithm is to switch these rules until the parameters converge.
The stability analysis consists of two steps. The first step shows that the parameter adjustment rules are switched at most finite times. In the second step it is shown that all the states of the control system are uniformly bounded and the output error converges to zero.
Computer simulations are performed to illustrate the validity of the proposed control system.

抄録全体を表示

The way to control master-slave manipulators affects considerably to the maneuverability of the master-slave systems. The ideal state of master-slave systems can be regarded that the operator can operate the system as if he were directly manipulating the object which is actually existing at the remote site. In other saying, the system must be coupled with the operator to give the ideal kinesthetic sense so that he can perceive the object. So far, several researches discussed about the ideal states of master-slave systems in their own descriptions. However, there is few exact discussion about how close the ideal state can be achieved actually or what kind of control scheme should be designed in order to achieve it. In this paper, we propose a control scheme which can achieve the ideal kinesthetic coupling with the operator and realize three ideal responses which were previously defined by the authors. Secondly, we show the stability of the system controlled by the proposed scheme by using the concept of passivity. We then discuss about the system stability when the sensor signals are pass through the filters. Lastly, the validity of the proposed scheme is confirmed by simulations.

抄録全体を表示

The identification of time delays in continuous-time systems is considered, and a simple recursive algorithm is derived with the use of Newton's method. Consistency properties of the algorithm are analyzed using the ordinary differential equation approach. It is shown that the algorithm leads to the true, unique solution if a ramp input is used. The algorithm is also applicable to the identification of time-varying delays.

抄録全体を表示

This paper proposes control methods for a fast-driving vehicle by fuzzy reasoning. The speed and steering angle of the vehicle are controlled by fuzzy controllers according to the distance and direction of one or two target points. The two target points tracking control requires more fuzzy rules and external informations than the one target point tracking control. However, the two target points control causes smoother steering, acceleration and motion than the one target point control.
A vehicle model has one rear wheel drive and one front wheel steering, and is described by non-linear differential equations in consideration of the dynamics. Using this model, the computer simulations of vehicle motion are carried out to show the effectiveness of the proposed control methods.

抄録全体を表示

In constrained tasks such as deburring, polishing and assisting disable persons, not only a position but also a contact force must be controlled. In addition, a reasonable compliance is necessary to absorb excessive impact or contact forces. A pneumatic robot can well meet these requirements because of its inherent compliance due to air compressibility. In addition to this passive compliance, an active compliance control method of a pneumatic robot is considered.
A compliance control system regulates the relation between the manipulator position and the contact force along each task axis. This study proposes a position-based type compliance control system which has a minor position control loop inside a force feedback loop. This position control loop dominates the overall compliance control performance. Unfortunately, a pneumatic robot can not easily provide the high position control performance with usual control schemes owing to the low stiffness and the high nonlinearities. To overcome this problem, a sliding mode control scheme with a pressure feedback compensation is adopted to the position control loop. This compensation is effective to inhibit the chattering.
The proposed compliance control method is applied to a pneumatic robot comprising two arms. First, both stability and control accuracy are examined theoretically and experimentally. Next, this method is applied to some tracking controls on the constrained wall. Further, a stabilizing method to realize the higher compliance is proposed and its availability is confirmed. The results obtained assures the effectiveness of the proposed control method for a pneumatic robot. Consequently, the availability of a pneumatic robot may be increased.

抄録全体を表示

The present paper proposes a multi-point compliance control for manipulators whose end-points are constrained by task objects. Compliance of the contact point is affected by the mechanical compliance of the objects as well as the joint compliance of the manipulator. Therefore, to derive the joint compliance which realizes the desired contact point compliance, the manipulator and the task object need to be represented as the parallel link structure. The method presented here can regulate not only the contact point compliance but also the compliance of several points on the manipulator's links utilizing kinematic redundancy.
First of all, the manipulator and the object are divided into a couple of virtual arms whose end-points are connected each other at the contact point or the points on the manipulator's links. Consequently, the kinematic structure of the manipulator becomes the parallel link structure and the relationship between the stiffness of the connected points and the joint stiffness of the manipulator is formalized. Then the optimal joint stiffness is derived, which realizes the desired connected point stiffness as nearly as possible. Finally, it is shown that the multi-point compliance control can regulate the compliance of the connected points for obstacle avoidance while controlling the contact point compliance.

抄録全体を表示

This paper proposes an alternative coordinated motion control architecture of robot arms manipulating an object. The motion and the internal force of the object are resolved into the motion of each arm. And each arm is controlled based on the virtual internal model so as to operate in coordination even if geometric errors exist in the robot arms and the object. The virtual internal model is a reference model driven by sensory information implemented in the controller. The proposed architecture will keep the stability of the system even if the breakage of the manipulated object occurs. The control algorithm is experimentally applied to the coordinated motion control of two planar robot arms, each of which has three degress of freedom. The results illustrate the validity of the proposed control architecture.

抄録全体を表示

The problems of recent approaches in qualitative and causal reasoning include; (1) In qualitative modeling, they do not discuss on what types of mathematical models their qualitative equations should be based. This causes them to misuse static models for the qualitative equations to discuss changes in the transient state. (2) Many of them use causality which does not refer to time explicitly. Then, in reasoning, they substitute logical inducibility or mathematical derivability for the causality. Such causality may include undesirable characteristics for producing causal accounts for physical systems.
In this paper, we define the “inherent causality” which requires “time reference”. With this causality, causal reasoning is carried out by verifying that any change is made by the causality. The verification is done by the requirement of dt=+ for each step. In order to make such causal reasoning possible, we carefully choose the base model, i.e. the dynamical model which describes causality from what makes the change to what is changed. We also developed the qualitative simulation algorithm with the inherent causality. The power of the causality and simulation algorithm is demonstrated on two examples of the pressure regulator and the mass-spring system.

抄録全体を表示

Ocean thermal energy conversion (OTEC) is a system that converts heat energy into electricity by using the temperature difference between warm surface sea water and cold deep sea water. Computer simulation plays an important role in designing and controlling the system. It helps to grasp precisely the characteristics of the system and therefore, enhance an effective and steady operation of the system.
This paper describes the computer simulation for a generating system based on OTEC plant. To construct this computer simulation, the system was broken down into three parts; Evaporator, Condenser and Turbine, and mathematical equation models were obtained for earth parts by using physical theories and reflecting properly the structual feature. Through these mathematical equations, we constructed a computer simulation for the OTEC plant. By using this computer simulation, we examined the operating characteristics of the plant for the following cases; (1) A drastic change in sea water temperature. (2) A drastic change in sea water flow rate and working fluid flow rate. We then compared the experimental result with the computer simulation result obtained. We confirmed that this computer simulation expresses the dynamic characteristics well.
The results obtained from this computer simulation will be used effectively to determine the plant specifications and to evaluate the economic prospectives of the OTEC plant.

抄録全体を表示

This paper is concerned with a new type of optimization problems, which are called “sequentially constructed problems”. This type of problem is not completely well-defined at a point of time when a optimization algorithm starts, and is sequentially defined with the progress of the iterative algorithm. Such problems arise in an optimization approach integrated parameter identification by observation of system input/output signals.
Currently used algorithm of mathematical programing, such a descent method, is unsuitable to solution to these sequentially constructed problems, because permanence of the problem is prerequisite when the algorithm is performed. In a case when the usual method is applied unconditionally, the convergence and stability of a sequence of trial points obtained by the algorithm cannot be assured.
In this paper, an algorithm for this new type of optimization problems is proposed, and the sufficient conditions for stability and convergence of this algorithm are also given.

抄録全体を表示

This paper describes an efficient method that finds collision-free paths for a multiple-degree-of-freedom manipulator with rotational joints and grasped object. The method first analyzes the structure of empty space in the 3-D workspace. Based on this space analysis, the path search is divided and a most promising direction for path search is determined in the 3-D workspace. Finally the path search is done systematically in the joint space in the direction equivalent to the promising direction. This method is applicable to various problems regardless of the number of degrees of freedom of the manipulator, its structure, and the presence of grasped object.

抄録全体を表示