We have proposed an automated image registration method ART for multi-temporal processing between remotely sensed multi-spectral images. ART had a binary triangle division process, in which triangle was recurcively divided into two sub-triangles so that discrepancy in it was removed by a linear transformation. ART achived highly efficient processing in registration of remotely sensed images. ART, however, had redundancy in triangle division. When a new control point P₁ was selected on a common side between triangles T₁ and T₂, the corresponding point was searched twice for division of T₁ and T₂, respectively. As the searching was done independently using an index of spatial correlation in the triangles, selected corresponding points were not always identical. Due to the redandancy, one to one correspondence between the reference image and the registered one was not satisfied. Gap or overlap between neighboring triangles decreased the accuracy of the registration.
In this paper, we propose an improved method TART (Improved ART) to realize both high accuracy and high efficiency. For reducing the redundancy in the triangle division, TART uses a forced division in the process. The forced division removes the gap or overlap and yields more triangles without the redundant searching process. The validity of TART is confirmed by a numerical simulation and by processing actual remotely sensed images. Robustness of IART for miss-selection of initial GCP pairs is also shown.

抄録全体を表示

This paper describes about haptic aspects and these graph representation of tactual images.
As the acquisitions of tactual images are carried out by the tactile sensings onto an object's surface, it is necessary to provide with the representation of the tactual images due to the geometric relationships with the features of an object's surface.
The haptic aspects are defined as the equivalence classes corresponding to the tactual images which are classified by the equivalence relations to image set acquired by the tactile sensings with any exploring directions onto a given object's features.
The haptic aspect graph representation is done as a graph for the geometric relationships of the haptic aspects which are classified with the tactual images for the features of the object's shape; edges, faces, vertices, and concave features.
A high density optical tactile sensor has been developed for acquiring the tactual images from the object's surface. The pose control of the object's face are confirmed by usingg the normal vector of the face which is estimated from the tactual images acquired by the optical tactile sensor.

抄録全体を表示

This paper presents a telemetry capsule for measuring gastric pH. The electrode of the new capsule is iridium oxide. Contrary to an antimony electrode, this iridium electrode does not possess toxicity and is more solid than a glass electrode. The electromotive force and the diameter of the iridium electrode is 57mV/pH and 0.25mm. The process of modulation is PFM-FM, because the instability of the frequency of the FM circuit lacks reliability to pH monitoring using telemetry. The PFM circuit is a voltage controlled oscillator (VCO) circuit. The current consumption of VCO is 770μA. The range of the frequency variation is less than ±10Hz when the center frequency of VCO is 6.54kHz. The voltage of the power supply which is two 1.5V silver oxide batteries is held in 2.5V by two voltage regulators. The electronic circuits required for the capsule were customized into 2 hybrid ICs. The capsule diameter and length are 12mm and 28.5mm. The total current consumption of the electronic circuit is 3.51mA. The measurable time is 2.5 hours and the error is 0.2 pH. In order to confirm the usableness of our capsule, the gastric pH of a fasted dog was monitored. After confirming that the capsule was in the stomach with an endoscope, ranitidine which is and inhibitor of acid secretion was injected to the dog's vein in order to measure pH alteration. After intravenous injection, pH value started to increase from pH1.5 to pH4 remarkably. Though the suction pump and the driver of the endoscope were working during this measurement, the demodulated wave form was not influenced by the electromagnetic noise of these devices. We conclude that our capsule is usable to measure gastric pH. Further studies, however, are necessary for improvement of the miniaturization of capsule size for clinical application.

抄録全体を表示

Most algorithms currently proposed for Sensor Fusion assume that the data from multiple variant sensors are coherent (synchronus and without time delay of measurement). In this paper, we propose a new Sensor Fusion algorithm which fuses incoherent multi-sensor data according to the fixed sequence of time-variant combinations of sensors. This algorithm is proved by the existence of equivalent Kalman filter, which corresponds to this sequence. observability and controllability of equivalent Kalman filter assure the steady state solution of unique estimation error covarince matrix. We also propose a new method of evaluating Sensor Fusion system by mutual information rate.

抄録全体を表示

Piano and guitar tone decay analysis has been carried out using reverberation responses calculated by Wigner distributions. The responses are investigated in comparison with record of the conventional short time spectra. The reverberation responses which are intimately related with the internal energy of instrumens represent 1) the envelope of tone decay without fluctuation induced by interference, reflect 2) mutual load effect associated with the oscillators, thus, clearly show 3) the gradients of energy decay curves, and are 4) less influenced by window function length. A guitar under investigation is shown to provide 1) two-step tone decay curve not caused by difference between decay rates of horizontal and vertical vibration modes of the strings (single string tone), and 2) multi-step energy curve with each step having much longer duration than the interference period (two string tone).
On the other hand, the analysis of a grand piano reveals the following features:
1) two power peaks in the spectra of aftersound are caused by different reasons from each other, 2) multistep energy decay curve is altered by performance style or piano condition, and 3) the partials are exhausted after absorbing energy of the other vibration modes.

抄録全体を表示

This paper is concerned with the first novel massflow measurement of stand pipe for moved bed gravity flow. Aim of this study is the confirmation of a novel massflow measurement methods for gravity flow in stand pipe in the new process that is high calorie coal gasifier or direct reduction process. The gravity flow in stand pipe is required for high pressured, high temperatured and corrosive feed conditions. This paper describes heat transfer method based on one of the author's efforts for pulverized coal injection system. It is required for reducing accuracy. It is clearized by the authors that the wall temperature is the most important for the system and there is a boundary layer heat transfer on the inside surface of the pipe.
Scale up of meter size and expansion of feed specification will be studied in the next stage.

抄録全体を表示

In this research, the relation between passive systems and finite gain systems are considered, and in the view point of this the equivalence between solving nonlinear H^∞ problems and finding the feedback which renders a nonlinear system passive is shown. First the state space representation of Cayley-transform (H-I) (H+I)^-1 of a nonlinear operator H and that of its inverse transformation are proposed. Next Cayley-transform is extended to the case of general control systems.

抄録全体を表示

In recent years, a robust stabilization problem in the gap-metric has received much attentions and, for finite-dimensional systems, it is recognized that the design procedure enables us to take a trade-off between the robustness and the performance of the closed loop system. Moreover, this argument has a merit to obtain a control law in a closed formula, which has a well-known feature of LQG-controllers.
In this paper, we focus on a system with delays in control and consider the robust stabilization problem in the gap-metric. By employing completing the square argument of particular quadratic forms, a design procedure of robust controllers is derived in the framework of finite-dimensional operations. The control law is constructively given based on the solutions to matrix Riccati equations and has a feature of observer-based predictive action. Then we clarify that an alternative representation of maximal robustness margin is characterized with a solution to an transcendental equation.
A game theoretic interpretation is also provided on the trade-off between the initial-uncertainties and attenuating the disturbance caused by plant uncertainties.

抄録全体を表示

This paper is concerned with robust decentralized control systems.
This paper presents a robust stability condition for decentralized compensators to stabilize the given plant with coprime factor perturbation. A robust performance condition for decentralized compensators is also presented. This condition gives constraints on individual decentralized compensators. And it enables local compensators to be designed independently.

抄録全体を表示

In this paper, the existence of realization in a linear time-varying discrete-time system is considered. First, the notion of ‘semilinear transformation’ for a time-varying discrete-time system is introduced. We study necessary and sufficient conditions for the unit pulse response to be realizable by a q-step observable system or r-step reachable system. These conditions are also represented in terms of Hankel matrices.

抄録全体を表示

This paper studies a control design approach for certain linear discrete-time systems of single input and single output where an assumption of minimum phase systems and detecting correct delay-time are not required. First, we derive a condition for the optimum finite settling trajectory of the system. Then, we find that a control to transfer output and input value of the system to the optimum finite settling trajectories, which are successively predicted at every sampling time, is obtained in an output feedback form which is derived from the above condition. We verify that all control systems designed on the basis of the design method proposed in this paper are asymptotically stable without requiring any condition of systems to be controlled. This design method can be immediately extended to adaptive control by adding a suitable on-line model identifier. This adaptive control scheme is applicable to chemical plants which have uncertain and changing large time lags.

抄録全体を表示

This paper proposes a digital preview control method for nonlinear mechanical systems. On the trajectory control of mechanical systems, the preview control method utilizing future information of the desired trajectory is effective to improve the trajectory control performance. In the meantime, on the nonlinear mechanical systems control it is important to compensate the nonlineary dynamics. In the previous paper, we proposed a digital preview control method that could realize both nonlinear dynamics compensation and trajectory control from a viewpoint of the motion process. However, it was a practical problem that the previous method required to sense the acceleration of mechanical systems. In this paper, we primarily derive velocity and position state equation as a discrete-time model of mechanical systems based on the motion process. And then we design a digital preview control system based on the model. By means of this procedure, sensing the acceleration becomes unnecessary. Finally, the effectiveness of the proposed method is verified by simulation studies and experimental studies.

抄録全体を表示

An identification algorithm that enables frequency domain weighting to reduce influences of unmodelled dynamics or noises is proposed. In this paper followings are shown using some examples: 1) It may happen that frequency bands to be empasized, in which signal-to-noise ratio is high or in which model fitting accuracy are needed, are known from some informations about plants, sensors or specifications for control. And this method is successfully used to weights such frequency bands. 2) The proposed method requires the smaller degree identification models to achieve sufficient data fittings compared to the well known least-squares method. 3) The computational power required to execute the proposed algorithm is not so big for popularly used work stations.

抄録全体を表示

In the many magnetic disk drives, several servo controllers are switched from one to another to perform one servo action. This type of servo controller is called a mode switching controller and obtains fast access and accurate positioning. The authors previously proposed a basic idea of applying initial value compensation method to the mode switching controller at switching in order to improve transient characteristics.
In this paper, an actual design method for applying to a head positioning servo of magnetic disk drives are shown. Zeros of the transfer function between state variables of the plant and a controlled variable (head position) can be relocated by injecting a suitable values to the state variables of the compensator at mode switching. These values are calculated by using detected plant state variables at switching. Therefore the relocated zeros can cancel some undesirable poles and this improves transient characteristics with no influence on a closed loop stability. In the case that the detected plant variables have some errors, a complete pole-zero cancellation is not obtained. In the practical way of avoiding this problem, a specified servo mode is added between fast access mode and accurate positioning mode, whose poles are assigned not to include slow modes. By combining this mode with the initial value compensation method, fast transient with less overshoot can be obtained. Finally some experimental results are shown.

抄録全体を表示

To achieve the safety of intelligent systems for real-world applications, not only processor faults but also real-world errors such as sensor errors, failure of the controlled object and operation errors must be covered. From this point of view, an intelligent fault-tolerant system for real-world applications is proposed based on triple-modular redundancy. The system consists of a master processor that performs the actual control operations and two redundant modules. In the redundant modules, real-world process is simulated together with the control operations using knowledge-based inference strategy. To keep the independency between the triplicated modules, the simulation for error detection and recovery is performed without actual external sensor signals used in the master processor. It is made clear that the safety operation can be assured for a very wide ranges of faults and errors such as sensor errors, failure of the controlled object and operation errors.

抄録全体を表示

In this paper, we propose an algorithm for estimating position and orientation of an object grasped by a robot hand. The number of vertexes required to estimate the orientation is clarified by adopting elements of a rotation matrix as parameters of the orientation. The estimation process is as follows. The estimation problem is formulated as a problem of getting a rotation matrix that minimizes the error norm. A maximization problem of a function with two variables is derived from the minimization problem. It is difficult to solve the maximization problem analytically, but the function has unimodality. Hence, the maximization problem can be easily computed using a direct detection method. And initial estimated values are not needed.

抄録全体を表示

This paper describes a method of automatic detection of faulty holes of a shadow mask which is used in a cathode-ray tube of a color television. The image data are taken from two areas of the mask with CCD camera, and comparing the shape of holes in these two areas by use of correlation technique, we can find any fault in the shape of holes. Experiments with a shadow mask have shown that this method of fault detection is very effective for practical use.

抄録全体を表示

We describe a simple method for the simultaneous determination of burning velocity and quenching distance in laminar flames from measurement of flame propagation speeds. The method is applied to determine these two parameters in propane/air laminar flames at different equivalence ratios. The maximum burning velocity obtained using our method is higher than any previously reported value. The theory behind the new approach is also described.

抄録全体を表示

This paper will describe the new nondestructive measurement of pesticide residues in the agricultural products. This method proposed uses Fourier transform infrared spectroscopy of attenuated total reflectance method (FT-IR-ATR). This method processes the spectrum by the canonical analysis and or quadratic discriminant analysis. It takes about fifteen minutes to measure the pesticide residues in the agricultural products. The measuring time is the shortest among the conventional destructive methods, of which measuring time are some hours. This method was applied to some kinds of vegetables that were practically produced in the real fields. The discrimination rules were verified using the Gaschromatograph values as the references.
The main results of this study are as follows.
(1) The discrimination rules are proved to be stable in the case of lettuces as the test set. The correctness of discrimination is around 80%, that is as much as one in the case of the training set.
(2) The discrimination rules are expected to be applicable to the other kinds of vegetables; celeries and chinese cabbages, while they are not applicable to cabbages, because of its surface properties.
(3) The safety inspection system of the agricultural products using the FT-IR-ATR are expected to be more effective and lower cost than the conventional methods.

抄録全体を表示