This paper is discussed the trend of technology to be achieved the low phase noise in the oscillator in recent years. The oscillator is generally designed to satisfy the oscillation condition, as the gain of the amplifier more than the loss of the feedback circuit. However, we have not understood which type of configuration of oscillator circuit is better for the extremely low phase noise.
In this paper, we have discussed on a configuration of oscillator circuit to obtain the extremely low phase noise, as refer to an idea of Noise Figure in the amplifier circuits. We have proposed an oscillator circuit operating at a non-reactive frequency of SAW resonator, called as Fr oscillator, in a point of view as to be achieved an extremely low phase noise. Moreover, we have discussed on the signal processing technique to obtain the extremely low phase noise.

抄録全体を表示

One of the most serious problems in recent VLSI systems is data transfer bottleneck between memories and processing elements. To solve the problem, a model of highly parallel VLSI processors for intelligent integrated systems is presented. A logic-in-memory module composed of a processing element, a register and a local memory is defined as a basic building block to form a regular parallel structure. The data transfer between adjacent modules are done simply in a single clock period by a shift-register chain.
A high-level synthesis method is discussed on the hardware model, when a data-dependency graph corresponding to a processing algorithm is given. We must simultaneously consider both scheduling and allocation for the time optimization problem under a constraint of an chip area. That is, we consider the best scheduling together with allocation such that the processing time becomes minimum under a constraint of a fixed number of modules. Not only an exhaustive enumeration method but also a branch-and-bound method is proposed for the problem. As a result, it is made clear that the proposed high-level synthesis method is very effective to design special-purpose VLSI processors free from data transfer bottleneck.

抄録全体を表示

A differential optical absorption spectroscopy (DOAS) system for measurement of atmospheric NO₂ was developed. The system uses a battery-operated, high luminance LED and a fiber-coupled spectrometer, and is portable. Laboratory experiments using a gas cell of length 0.22 m with varying NO₂ concentrations were performed to evaluate the sensitivity of the DOAS system. The DOAS measurement results are in agreement with NO₂ concentrations obtained simultaneously by a FT-IR (Fourier Transform Infrared) system for NO₂ concentrations down to 20 ppm. Experiments with an optical path length of 93 m were also performed, and NO₂ concentrations down to 0.20 ppm were measured. Since measurement of atmospheric NO₂, which is in the order of several tens of ppb, requires optical path lengths of several hundred m, system improvements to improve the signal detection are necessary.

抄録全体を表示

Left ventricular thickness, diameter and their time variance play important role in an assessment of a cardiac function using echocardiography. In order to measure left ventricular wall thickness and diameter from echocardiogram, it is required to determine a center of left ventricle and to detect a border which is a boundary between cardiac muscle and blood.
On the other hand, the center of left ventricule is regarded as the varycenter of the border curve of left ventricule, but the varycenter is able to be determined after detection of the border curve.
In this paper, we propose the new method to detect the border curve and determine the center of left ventricle as the varycenter, at the same time, shifting the origin of coordinate system.
Furthermore, this method realizes about twice high-speed compared with a conventional method.

抄録全体を表示

The brain hypothermia treatment is an attractive therapy for the neurologist because of its neuroprotection in hypoxic-ischemic encephalopathy patients. The present paper deals with the possibility of controlling the brain and other viscera in different temperatures from the viewpoint of system control. It is theoretically attempted to realize the special brain hypothermia treatment to cool only the head but to warm the body by using the simple apparatus such as the cooling cap, muffler and warming blanket. For this purpose, a biothermal system concerning the temperature difference between the brain and the other thoracico-abdominal viscus is synthesized from the biothermal model of hypothermic patient. The output controllability and the asymptotic stability of the system are examined on the basis of its structure. Then, the maximum temperature difference to be realized is shown dependent on the temperature range of the apparatus and also on the maximum gain determined from the coefficient matrices A, B and C of the biothermal system. Its theoretical analysis shows the realization of difference of about 2.5°C, if there is absolutely no constraint of the temperatures of the cooling cap, muffler and blanket. It is, however, physically unavailable. Those are shown by simulation example of the optimal brain temperature regulation using a standard adult database. It is thus concluded that the surface cooling and warming apparatus do no make it possible to realize the special brain hypothermia treatment, because the brain temperature cannot be cooled lower than those of other viscera in an appropriate temperature environment. This study shows that the ever-proposed good method of clinical treatment is in principle impossible in the actual brain hypothermia treatment.

抄録全体を表示

In the observation through the slit, we can observe only the part of the object in some moments. However, we can perceive the whole image, when the object moves behind the slit. Then, it is perceived to be shorter than the length of actual object. This causes “Anorthoscopic perception.” This phenomenon indicates the mechanism for perceiving the whole image which temporarily stores the fragments of the image. The mechanism is not yet investigated in detail. In this paper, we propose the technique in which we quantitatively measure the shortening effect using multiple slits. We carried out the psychology experiment using this technique. Then, we obtained the result which indicated the possibility with the relation between the effect and eye movements. So, we measured the eye movements in observing through slits for verification of this possibility by the EOG-method. We are focuced on the delay and generation rate of smooth pursuit eye movement.As a considering this factor, we predicted the shortening rate. The predicted curve was matched experimental result.Finally, it was indicated that the delay and generation rate of smooth pursuit eye movement played important role in the shortening effect of an anorthoscopic perception.

抄録全体を表示

Data of scattering coefficient on vertical incidence against ground surface at U-band are obtained by Millimeter Wave Radar Altimeter using FM-CW ranging. Noise suppression by wavelet shrinkage can be utilized to extract feature parameter in high spatial frequency band, in which level of fractal noise is dominated by that of white noise. We propose approximate algorithm for estimation of local fractal dimension in high spatial frequency band which is the most effective parameter for identification for classification of ground surfaces such as vegetation, town area and rice field.

抄録全体を表示

We propose a new multi-level error diffusion method employing the variable threshold, which realizes half toning with high quality. In the proposed method, dissimilar textures around quantization levels, which are the cause of the low image quality and obtained by a conventional method, are decreased by using the variable threshold. In this paper, we apply the proposed method to various standard digital images. And the simulation results show that the method is superior to other multi-level error diffusion methods.

抄録全体を表示

This paper describes a robust tracking method that combines a scanning laser radar (SLR) with a camera to measure headway distances and lateral positions more accurately. With this method, a preceding vehicle is first detected on the basis of the digital data output by the SLR unit, and the vehicle is then tracked by both SLR and the camera. A variable shape template matching method is used in processing the images taken with the camera. The image recognition technique can track the preceding vehicle despite frequent changes in the headway distance and the direction. Robust and highly accurate tracking is achieved by combining the wide angle, high-resolution measurement obtained by image recognition with the SLR distance measurement. The proposed method has been validated on a driving simulator using data recorded on Japanese roads. The preceding vehicle was successfully tracked and its position was measured as expected on curves and on straight roads in general. The results also confirmed that this method has some advantages over detection by SLR alone in difficult detection situations.

抄録全体を表示

A new design method of digital preview sliding mode servo system is proposed. In this method, Sliding mode regulator theory is applied to an augmented error system. The equivalent control which make the states of error system stay on the switching surface is employed. Then, the sliding mode controller is designed. Preview sliding mode servo system by using this control has both merits of the preview control and the sliding mode control. The computer experimental results on LDM position control systems show the effectiveness of the proposed design method.

抄録全体を表示

This paper describes a method of a network flow analysis under water shortage and water distribution control with reduced pressure. As a water shortage measures, the water supply pressure should be kept at lower level than usual to save water. In this status the control of pumps and valves are difficult because the demand varies according to the pressure. In a conventional method of the analysis the pressure at the water supply bases and the demand distribution on the pipeline network are given. Then, the entire pressure and flow distribution can be obtained by solving the equations of the pressure equilibrium and the flow quantity preservation. When the pressure is not sufficient, the demand at the node is the function of the supply pressure and the usual method does not apply.
We enhance the conventional network flow analysis by introducing the relation between the supply pressure and the volume of water taken out of the node. A control method based on this analysis is proposed with which the pumps and the valves are controlled correctly. A simulation using actual data shows that adequate operation can keep the water saving target while maintaining uniform pressure distribution as possible.

抄録全体を表示

This work considers an adaptive servosystem design for a class of nonlinear systems that can be transformed into a canonical form. Although our proposed compensator is based on nonlinear internal model principle, overall control system structure adopts plug-in manner, i.e. the compensator to achieve tracking and disturbance rejection is placed outside the existed feedback controller. Moreover such compensator can be designed by solving error feedback nonlinear regulation problem and by using recursive procedure. An adaptation mechanism is introduced to cope with the unknown parameters of exosystem.

抄録全体を表示

Complex neural networks (C.N.N.) with their complex back-propagation learning algorithm are useful in processing complex signals. As the characteristics of the C.N.N., the learning and generalization abilities with respect to affine transformation on a complex space are known. This paper is concerned with applications of these C.N.N.’s characteristics to generation of fractal images by constructing iterated function systems (IFS) with the C.N.N.’s input and output relations. The simulations for several examples demonstrate effectiveness of the C.N.N. to coding fractal images in the prospect of designing the desired fractal images.

抄録全体を表示

Pervasive computing is becoming a more important issue for open distributed systems. It is, however, hard to design such distributed systems when the user uses many kinds of terminals and migrates with these. This paper has proposed a new framework for pervasive computing. The framework includes three kinds of agents: User Interface Agents, Programmable Agents and Service Mediation Agents. We can easily design a flexible distributed system by collaborating these agents. We also evaluate the framework from flexibility point of view.

抄録全体を表示

Although there have been many studies on parallel machine scheduling problems, the number of machines operated is fixed in these studies. It is desirable to generate a schedule with fewer machines operated from the viewpoint of the operation cost of machines. In this paper, we cope with a problem of minimizing the number of parallel machines subject to the constraint that the total tardiness is not greater than the value given in advance. For this problem, we introduce a local search method in which the number of machines operated is changed efficiently and appropriately in a short time as well as reducing the total tardiness.

抄録全体を表示

In this paper, we proposed a procedure for calculation of room temperature in multi-room building using the backward-difference method. Based on this procedure and using our simulation program, we evaluated CO₂ reduction by installating high light-reflective and high heat-emissive paint in buildings. Heat balance equations on each heat point in each room or wall in the building were deduced from one-dimensional heat-conduction equation and converted to the matrix equation. We improved matrix computing and developed a different procedure from conventional methods. Because this procedure is simple, we consider that this is applicable to estimation of most building’s technologies. As its application, the effects of high light-reflective and high heat-emissive paint was evaluated. Although the paint increases heating load, it can reduce cooling load. We thus conclude that the paint is effective for CO₂ reduction.

抄録全体を表示

Due to the spread of distributed systems, a number of agent systems have been developed. Since most existing multiagent systems implement different agent platforms, it is extremely hard to get heterogeneous agents to work together. In this paper, we introduce a new network protocol called Agent Platform Protocol (APP), which is designed to meet the exact demands of agent interaction over world-wide networks. The most significant contribution of this protocol is that it supports Peer-to-Peer communication among agent platforms and hides the network-level implementation details from agents. To demonstrate its usefulness, APP has been used to realize the interaction between independent agent systems from two projects, JADE and KODAMA. JADE is an agent system in compliance with the FIPA specifications, and KODAMA is an agent system one being developed in our laboratory. The experimental results show that APP can help JADE and KODAMA agents to communicate with each other without getting the agents being aware of the difference in their agent platforms.

抄録全体を表示

We have researched on consultation of feeling of Pleasantness and Arousal by analayzing EEG for sensitive communication support. In this report, we suggest a method of visualizing them on real time.

抄録全体を表示

Blind convolution and deconvolution are fundamental signal processing techniques in acoustics. Recently, many blind techniques have been proposed without measuring the impulse response. The purpose of this study is to evaluate blind envelope estimations involving the carrier modeling of the source signal and room impulse response. We compare the envelope convolution method and power envelope convolution method on an experimental basis. The result of the envelope estimation obtained by the power envelope convolution method is better than that obained by the other method. In the reverberant envelope estimation experiment, the envelope distortion can be reduced by about 6 dB.

抄録全体を表示