The Cadmium Zinc Telluride (CdZnTe) semiconductor is recently utilized as a detector for the measurement of X-ray spectrum. In this paper, we calculate the energy absorption response of a CdZnTe semiconductor to mono-energetic photon beams by means of the Monte Carlo simulation, and investigate their energy dependence property. The photo-peak efficiency, which represents the detectable efficiency, and the fraction of K-escape showed three discontinuous points at same photon energies corresponding to the K-absorption edge of CdZnTe. X-ray spectra measured by a CdZnTe detector are different from the incident X-ray spectra, because of distortions caused by energy-absorption response of the detector. The net incident X-ray spectrum was well estimated by the stripping correction method using the response data of the detector to mono-energetic photon beams.
Fast and flexible encryption technique are required under the rapid spreading internet environment. Encryption technique generally consume the computational power and need specific hardware for practical uses. Specific hardwares can't be modified when the encryption algorithms are improved. In this paper, we propose the implementation of Reconfigurable Computing (RC) technique for encryption processing. RC is capable of accelerating information processing using dynamic reconfiguration of Field Programmable Gate Arrays (FPGAs). By dividing the target problems into hardware and software processing appropriately, the computation time will become much faster. RC will enable us fast and flexible encryption processing. However, there are few systens to realize RC as a multi-purpose board. We implemented RC system onto FPGA board.
In this paper a design method of the robust PSS (Power system Stabilizer) is presented on the basis of the identified model. The basic idea is to partially identify each generator, which connects with a large-scale power system, as a single-machine infinity-bus system. Whereas the structure of mathematical model and parameters of each generator are regarded as already known, the only unknown parameters are equivalent reactance of transmission lines, the voltage and the phase angle of infinity bus system. The novel idea is that known and unknown parameters are coded in one system matrix, and the guessed values of unknown parameters are estimated by identification process. The controller is designed by applying H∞ control theory to the identified model and the obtained controller is reduced to the reasonable order. It is found that the time response of identified model closely resemble the simulated response, and the controller are almost not degraded by order reduction. Effectiveness of the proposed method is verified through some nonlinear simulations for the three machine system of radial type.
We proposed a micro robot which consisted of electromagnets and piezoelectric elements. The proposed micro robot, which has closed-loop nagnetic field, is operated on a magnetic floor. The principle of the micro robot is based on an inchworm motion. We describe the linear displacement of the micro robot on an inclined surface. The relationships among the electromagnetic force, the applied voltage to the piezoelectric elements, and the displacement are described. The rotational displacement is also described.
We propose a method for extracting rotating limb regions and estimating their motion parameters, which is based on curl of optical flow. Regarding optical flow under orthographic projection as a vector field, the proposed method computes the curl of the optical flow to extract rotating regions not depending on the position of the limb. We use the EM algorithm which can perform both of dividing the image into multiple moving regions and estimating motion parameters of each region. Since the proposed method doesn't require a simplicity of background, and can discriminate translationa movements produced by other moving objects in the background, it is applicable to a practical situation in which conventional methods don't work well. Experimental results with real image sequences demonstrate that the limb extraction can be performed well from noisy optical flow which are obtained from real images.
Due to the remarkable progress in computer network technology as well as artificial intelligence, in the very near future we Japanese may read English texts through CRT displays with the aid of Electronic English-Japanese dictionary. Whenever we encounter any unknown word, we may learn the sense from the dictionary by pointing the unknown word with a mouse. It is easy to display all the contents of the dictionary related to any word of concern. However, the contents would contain more than enough senses and the user may be at a loss which of them to choose. For example, if we look up the preposition “of” in Shogakukan Random House English-Japanese Dictionary, 2 nd Edition, we find 20 senses. Which of the 20 senses should be chosen in a given sentence? Language experts seem to make full use of various knowledge sources, such as syntax and semantics, to disambiguate the senses to narrow down to the right one. We propose how to implement some of language experts' knowledges in an intelligent dictionary system and conducted ‘paper simulation studies’ with promising results.
The aim of this paper is to obtain an exact 3 D model of a scene using multiple images from different camera positions. The reconstruction of 3 D shape using two images has problem such as being weak at noise. Therefore we present methods to reduce noise and to improve the accuracy of 3 D shape with multiple images. The system is divided into three stages: firstly, reconstruction of 3 D shapes at different camera positions, secondly fusing the 3 D shapes to obtain a para-ideal shape, and thirdly removing the outlier shapes and feature points by evaluation function, and fusing the rest of shapes. Even though the corruption of image data by noise is one of the unavoidable problems in any system, this paper shows how well the noise is removed by the proposed algorithms in multiple view points. We demonstrate a significant improvement of recovered shape. Experimental results show that our system performs well to remove noise with robustness. The maximum noise reduction rate is 82% in the real image experiment.
Transmission experiments for a local area network (LAN) using blue light emitting diodes (LEDs) with InGaN/GaN multiple quantum well (MQW) and plastic optical fibers (POFs) have been demonstrated. Audio analog signals have been transmitted upto 1800 kHz, which corresponds to optical 1.5 dB bandwidth of LED response. Response speed of these diodes has been investigated by changing operating conditions; DC bias, pulse amplitude and pulse shape. The LED rise time is reduced from 0.26 μs to 0.19 μs by changing the electrical pulse amplitude from 0.5 V to 1.2 V, respectively, whereas the fall time (0.20 μs) unchanges. On the other hand, the fall time is reduced from 0.20 μs to 0.14 μs by adding negative pulses after the applied positive pulses. It is made clear that direct-modulation speed of these diodes is limited by the time constant associated with device-capacitance. Electro-luminescence and photocurrent spectra of LEDs are also investigated and blue shifts have been observed for both peaks in the spectra.
Revised method for evaluating the insertion loss of electroabsorption modulators has been proposed and demonstrated using InGaAs/InAlAs multiple quantum well (MQW) modulators. Various losses such as propagation loss, coupling loss and residual loss, and their incident wavelength dependence are estimated quantitatively. As a result, propagation loss including optical confinement factor is 13.3 dB and 9.7 dB for incident light wavelength of 1531.5 nm and 1551.4 nm, respectively, for a 300 μm long MQW modulator with absorption band edge of 1510 nm. Modulation characteristics of MQW modulators for long-haul and high-bit-rate optical fiber transmission systems, as well as absorbed photocurrent spectra, and quantum-confined Stark effect, are also investigated.
Knitting is a kind of art work and it is not so easy for knitting beginners to construct a work that they would design. Thus knitting requires that designers have some artistic sense and technical knowledge. Until now we had already proposed the supporting system for knitting design. It can change simple designed figures into pattern-knitting diagrams by using changing rules, and it supplies designers complete CG images of knitting that they have designed. Generally designers use various colored strings to knit, however the system could deal with only monochromatic knitting patterns. In this paper, we have developed new version of the system and would propose a method for colored knitting design. The new version supplies the CG images of colored knitting design and can support many designers to use colored strings to knit.
The aim of this study was to develop a new exercise test using surface electromyogram (EMG). The eight young subjects were performed exercise using bicycle ergometer. The anaerobic threshold (AT) was estimated from gas exchange parameter (oxygen uptake and carbon dioxide output) and from EMG parameters (heart rate versus rmsEMG) using the V-slope method. The present study clearly demonstrates that AT estimates by gas exchange and by EMG have a high correlation. In addition, this method was applied to one elderly subject. As a result, appropriate AT was estimated as well as young subjects. In conclusion, the AT is abele to estimate by EMG measurement, and an appropriate exercise level can be easily obtained in this method.
Prevention of falling during walking or transfer is important for maintaining quality of life in the elderly. Analysis of fall among the elderly is useful not only for preventing fall, but also to find out disease in its early stage. The aim of this study is to develop an ambulatory monitor that records falling time and frequency for long period. The monitor consists of a fall sensor, analog and digital circuits including a microprocessor. A photo-interrupter is used as the fall sensor. The photo-interrupter outputs a trigger signal when a subject fall and then the microprocessor records the falling time. To minimize the power consumption, the photointerrupter is driven with pulse current, and the microprocessor that records falling time is only operated when the subject falls. The monitor operates continuously for one week. The monitor was evaluated for a nor-mal subject during daily activity and it operated without any trouble in total 120 hours. The monitor was also evaluated for nine hemiplegia patients during rehabilitation training and it detected a falling.
Handwritten words often have slant which need to be estimated and corrected before character segmentation and recognition. Several methods have been proposed for average slant estimation and correction. However, average slant estimation has the problem such that local slant will be overestimated or underestimated when the slant in a word varies from character to character. To solve the problem, this paper proposes three methods for local slant estimation, which are simple iterative method, high speed iterative method and 8-directional chain code method. The experimental results show that the proposed methods can estimate and correct local slant more accurately than the average slant correction.
X-ray in the nanometer electromagnetic region is very useful for nanospace fabrication, information and high-energy transmission and control techniques of Å-size structures in biomedical sciences and technology. In this paper, the electromagnetic fields in X-ray gradient fiber with random refractive index media are studied. Statistical mode conversion of beam propagation, beam broadening and beam center are investigated by stochastic process theory of electromagnetic field.
Near-field optical microscope is a new type of high-resolution microscope using evanescent fields. In nearfield optical microscope, power of the scattered light is measured as scattered evanescent light by the material using tapered micro-probes. The micro-probe is one of most important parts in the near-field optical microscopes. We have studied the light propagation through the optical fiber micro-probe. We have used Beam Propagation Method (BPM) to simulate the light propagation. A computer program is developed using BPM and we studied the fundamental characteristics of optical propagation through the fiber probe.
Physiological response during high-speed motorcycle (MC) operation was examined. Electrocardiogram R-R intervals (RRI) of 18 professional riders driving MC at the speed of 150, 200, and 220km/h were measured on a specific oval course. Using digital filters, four bands of RRI time series, 0.06_??_0.5, 0.15_??_0.5, 0.012_??_0.025, 0_??_0.012 Hz, which related to the stationary component of RRI, parasympathetic nervous systems, difficulty in driving, trend of RRI respectively were extracted. Changes in these parameters with driving speed were discussed. The analysis revealed a tendency for changes in the stationary components and high-frequency components to decrease as the driving speed increases. It was also confirmed that trend parameter was significantly different when the driving speed was 150km/h and when it was 200 or 220km/h.
As nerve impulse arrives at a skeletal muscle, the sarcoplasmic reticulum releases Ca2+ into muscle cells. Ca2+ concentration ([Ca2+]) increases, troponin conformation changes, and then ATP hydrolysis accelerates largely. Using the ATP hydrolysis energy the myosin heads stroke actin filaments, thus tension is developed. Recently, the stochastic behavior of biological molecules is revealed by single molecule measurements. In this article, using the above knowledge, tension development model at molecular level is developed. First, we propose a model in which a site binds legands, with stochastic state transition. Applying the ligands binding model to four of Ca2+ binding sites on troponin, the troponin activation phenomenon is demonstrated. Next, a model in which a myosin head binds ATP is constructed. Using the models, we estimated the occurrence probability of a myosin head stroke, Rstroke. It is shown that estimated Rstroke realizes the time course of muscle tension, of which peak is delayed and broader than that of [Ca2+].
In the lithographic processes of VLSI's, high setting accuracies between masks and wafer are required. The methods using moiré signals obtained by gratings were found to give very high alignment accuracies. In this paper, influences of the inclination of gratings for the alignment accuracies in modified and differential moiré systems have been investigated. The moiré signals and the alignment point have been calculated using computer simulation. As a result, it has been shown that the effects are small and the inclinations do not affect alignment accuracies in the moiré alignment systems.
Today, web-based collaborative learning is one of hot topics as computer-support education environment. Many researches/developments have focused on the functionality, interactiveness, multi-media authoring, organization of environments, etc. for the investigation of collaboration facilities, but not on the subjects for managing learning processes, grasping understanding levels of participants, promoting creative discussions and so on. In this paper, we address a collaboration-support method to manage learning processes, grasp learning progress of group, promote active discussions among participants and grow up creativeness of individuals. In order to attain this method we introduce a teacher-oriented mechanism. This mechanism simulates the behavior of teacher who organizes and coordinates the group learning in our real world: the teacher monitors learning progress and contents of discussions roughly, remarks well/bad-understanding students, advises effective and successful comments/hints timely if necessary. Our idea for constructing teacher-oriented mechanism is to provide two different means: resolution derivation scenario and answering path. The resolution derivation scenario is useful to detect and resolve impasse situation in which answering process does not progress, while the answering path is powerful to promote discussion of various methods for deriving answer.
Recently, subsurface radar is paid worldwide attention for detection of buried objects, such as gas and water pipes, archaeological remains and mines. For these applications in the underground media, subsurface radar of very high resolution is required. However, since the underground media is inhomogeneous, many undesirable scattered waves are generated and the received signals will have random fluctuations. In this paper, we present our statistical investigations on the effects of random media, such as fluctuation and distortion of target signal at the receiver, as a fundamental study of error in target detection using pulse radar. These analyses are essential to find the statistical relationships between the random media and target detection error. Simulation results are presented for various models of random media using 3-dimensional Finite-Difference Time-Domain (FDTD) method. The statistical evaluations for reasonable prediction of the above relationships are demonstrated by numerical examples.
The transient electromagnetic fields due to electrostatic discharge (ESD) between charged metals have wide-band frequency spectra up to the microwave region, which give a serious malfunction to high-tech information devices. For the above ESD fields, we previously analyzed them, using the finite-difference time-domain (FDTD) method, and showed that the metals enhance the field level according to the metal dimension. From the standpoint of reducing such the ESD fields, the electromagnetic fields caused by the spark between the metals with ferrite core attachments were investigated. The FDTD method was also used to compute the ESD fields. A FDTD algorithm for the magnetic field inside the ferrite core was newly derived. The results show that the cores attached near the spark gap considerably reduce the magnetic field level, which is also confirmed experimentally.
A far-field prediction from the common-mode current on the traces of printed circuit boards (PCBs) was investigated, which had partly been examined by Paul in the frequency range up to 200 MHz for electrically short traces on the PCB with a narrow bandwidth signal source. Two sample PCBs were prepared, which have two different lengths of balanced parallel traces that are not always electrically short. A battery powered signal generator excited each PCB so that we could expand the frequency range up to 1 GHz with sinusoidal waves. A surface current probe consisting of thin wires wounded on the middle of half-cut ferrite core was used to measure the trace currents. Calculated electric far-field strength based on the common-mode current was compared with the measured far-field strength at a distance of 3m for the above sample PCBs in the frequency range from 30 MHz to 1 GHz. Good agreement between the common-mode current predictions and the measurements was obtained.
Five color spaces, RGB, XYZ, YIQ, HSV and LUV color spaces, were examined to know which was the most suitable to extract a face region. From the observations of flesh-colored pixel distributions in those color spaces, a face region was most successfully extracted by an ellipse from the UV plane in the LUV color space. A proposed method using LUV color space and ellipse extraction was an extension of that proposed by Hongo et al. that used a circle instead of an ellipse, and it improved extracted results by 6 points in our agreement index. A real time flesh-colored region extraction system was constructed on a personal computer with a 400 MHz celeron CPU. It could extract flesh-colored regions with the rate of 12 frames, 160×120 pixels each, per second.
A new off-line writer recognition method was proposed. In the methed, a 2-D membership function was obtained by piling up characters as a dictionary. The recognition procedure was a multi-character matching in which 4 types of characters were used. This was an extension of our previous work that was also used 2-D membership function with α-cut. In this paper, the principal component analysis was used to characterize one's hand writing features. The indivisual dictionary was obtained from characters which were not far from the average character. The proposed system was applied to writer recognition of 108 students who wrote 46 types of “hiragana” characters, in total 24, 840 characters, and fairly good results, almost 100% recognition rates, were obtained.
We propose a movable finite automaton (MFA) models for describing the translation process based on biological findings, in which the sequence of nucleotides in the mRNA molecule is translated into a corresponding sequence of amino acids to produce a distinctive protein chain. We have constructed the MFA models of mRNA, tRNA contained the end factor, the amino acids, and the small and the large ribosomal subunits concerned with the translation process. Using these quasi-bioelements, an algorithm of the protein elongation is developed, and then the computer simulation is performed. Results obtained is discussed from the biological studies.
Nano manipulation technology using Scanning Probe Microscope such as Atomic Force Microscope (AFM) has been introduced recently. Nano world physics is different from the macro world such that the gravity can be ignored and the in fluences of the capillary force or the Van der Waals force are dominant. Furthermore, since the scale is different, the tip control in the z direction is important for not breaking the tip and samples. Depending on the experimental results, the visualization of the force value is important to control the z position. For reliable and easy nano manipulation, user interface is very crucial. The ideal interface characteristics for nano scale manipulation is discussed in this paper.
An automatic design method for power systems, to control electrical loads provided by customers and to ensure the power supply for the loads, is described. This method features (I) an electrical equipment model including the possible connections among each electrical equipment, (II) knowledge association networks to reduce maintenance cost of knowledge-base and to improve efficiency to select/specify electrical equipment, (III) a planning algorithm to find electrical path to control the electrical loads and to ensure the power supply, and (IV) the reactive re-design to automate to edit the power systems. APEC3 (Advanced Plant Engineering Cycle tool version 3) is an automatic design tool for power systems that supports electrical engineering processes in developing single line diagrams, specification balance statements (design rationales) and item lists. When engineers modify equipment specifications, APEC3 invokes the redesign function which automatically changes the specifications of all relevant equipment. To evaluate the effectiveness of the method, APEC3 was used in designing practical power systems of steel plants. The results showed that APEC3 can reduce the design time to 1/8 as compared to the usual design time. Therefore, the engineers can find optimal specifications of the power systems by evaluating a number of possibilities.
A new convenient method to estimate a motion pattern of knee joint center is proposed in this paper. Knee joint is not a point. The motion analysis of the joint is difficult, since the knee motion is constructed by sliding and rotation. The center of knee joint moves by knee motion. If a motion pattern of knee joint center is analyzed, knee joint trouble can be detected. Ellipse approximate small group analysis is developed in this study. In the method, knee joint center is measured by ellipse approximate of moving malleolus position. By using the analysis, the knee joint center is measured in low error rate. Using the method, trouble of knee joint will be predicted. In this study, malleolus position is measured by digital video image. Subjects have no muscle burden.
A workflow model is used to describe business processes when workflow management system is constructed. In the workflow management system, “deadlocks” means the interruption of workflow caused by mistakes of workflow model design. This paper proposes a method which detects all deadlocks in workflow model. In order to find out all deadlocks, the deadlock patterns which express deadlock by combinations of nodes are adopted. However, there is the problem that some deadlocks cannot be detected because of endless loop. In order to cope with the problem we define expanded deadlock pattern which can be applyed to the case in which former deadlock patterns cannot find out deadlocks.
Line detection using Hough transform is one of the robust image processing methods for noisy image. But Hough transform has a problem that the computation cost is very large. In order to ease this problem, many high-speed algorithms were proposed. Xu and Oja proposed RHT (Randomized Hough Transform) which reduces the computation cost by selecting the pair of the edge points at random. Independently, Kiryati proposed probabilistic Hough transform (PHT) to reduce the computation cost by voting a part of the edge points in the image to the parameter space. We proposed a new high-speed algorithm called RVHT (Randomized Voting Hough Transform) which combines RHT and PHT. And we discussed experimentally and theoretically about the computation cost and the performance of RVHT by comparing with the specially regulated RHT and the original RHT. Since the proof that RVHT is faster than RHT did not completed, we present the complete proof in this paper. And we discuss the detailed properties of the algorithm of RVHT combined the edge deletion methods, the edge selection methods and the parameter space initialization methods. We show the differences between these combinations in the experiments of RVHT line detection. The best result of RVHT is 14 times faster than RHT. We applied this RVHT to detect circles from the microscopic image of salad dressing for the taste quality control. We got the feasible results by using the algorithm of RVHT with over 3 times faster than algorithm using the normal Hough transform. And we will be able to use the Hough transform circle detection to the practical applications.
A method to design an electromagnetically coupled microstrip slot antenna is proposed. The method is based on an equivalent circuit approach, where the slot part of the antenna is analyzed by the FDTD method and the stub part is regarded as a transmission line. In this paper, the electromagnetically coupled slot antennae are designed using the equivalent circuit method in 2-10GHz. As a result, it is concluded that this method is applicable if the stub length is longer than the width of microstrip line. It is also shown that the method is suitable for designing an offset feeding slot antenna.
It was reported that a sparse coding algorithm produced a set of basis functions being spatially localized, oriented, and bandpass for natural images. The application of Independent Component Analysis (ICA) to the natural images has shown to be similar results to the sparse coding's result. However, the ICA can be applied in the case of basis function matrices to be non-singular and invertible. There are not such limitations in the sparse coding algorithm. This property allows that the code is overcomplete, that is, the number of code elements is greater than the effective dimensionality of the input space. The purpose of this paper is to examine what characteristics of speech the sparse coding algorithm extracts from natural sounds. Speech data was Japanese five vowels uttered by a female speaker during about 1sec. Most of the basis functions were localized in frequency after the training. Some basis functions only shifted in time and resembled each other. Each basis function was compared with the speech data and the result was that some basis functions responded selectively to each vowel. The frequency analysis for the basis function showed that some basis functions extracted the pitch frequency and the formant of each vowel.
Power Outage Scheduling Problem is a distributed constraint satisfaction problem where scattered local power stations have to make consistent schedules with one another. However, synchronous backtrack algorithms, a well-known method for distributed constraint satisfaction problems, have difficulty handling rapid schedule adjustments and impartial assignement of power station schedules. Thus we propose two kinds of heuristics: parallel assignment and multiple priority strategies. We also developed a distributed system for the power outage scheduling problem which makes use of the above heuristics in order to ensure efficiency. The system is based on Bee-gent (Bonding and Encapsulation Enhancement aGENT framework) and consists of schedulers for each power station and mediation agents which have cloning & merging functions to support the implementation of the heuristics. The result of the experiment shows the improvement when handling rapid adjustment and impartiality issues and that this improvement is provided by reasonable computational overhead.
From the viewpoints of safety operation and efficiency of transfer cranes, spreaders (suspended from the trolleys) are controlled to stop as quickly as possible when the trolleys are stopped. It is therefore important to measure the sway of the spreaders accurately. For this purpose, a CCD camera has been used which was attached on the bottom of the trolley. However, the accuracy of the measurement was insufficient because of the deflection of the camera axis from the vertical line. The paper thus first considers the real-time measurement of the deflection in the camera attitude (against the vertical line) based on the analysis of the dynamics of the girder system and strain gauges attached on the girder, and next presents a high-precision measurement of the sway of the spreader.
This paper presents a design of semiglobal robust tracking output feedback controller with almost diturbance decoupling for a class of nonlinear systems. The system is affected by uncertainty and unknown time-varying dicturbances. The uncertainty is described by gain bounded unknown nonlinear functions of the state variables. If the system is of relative degree one and of minimum phase, the proposed controller ensures input-to-state stability with respect to disturbance inputs, and almost disturbance decoupling properties for any initial condition and for any smooth bounded output reference signal.
We consider the backing up control of a vehicle with triple trailers via a model-based fuzzy control methodology. To begin with, the vehicle dynamics is represented by a Takagi-Sugeno fuzzy model which is equivalent to a simplified nonlinear model of the vehicle. Then we employ the so-called “parallel distributed compensation” design to arrive at a controller that guarantees the stability of the closed-loop system consisted of the fuzzy model and controller. The control design problem is cast in terms of linear matrix inequalities. The simulation and experimental results demonstrate that the fuzzy controller effectively achieves the backing up control of the vehicle with triple trailers.
Output predictions of Generalized Minimum Variance Control (GMVC) are obtained by solving the Diophantine Equation based on the present model parameters. Therefore, exact value of prediction would not be calculated for time-varying system changing model parameter fast. In this paper, we show output predictor and control law of GMVC on the basis of Generalized Predictive Control (GPC) for time-varying system suggested by O. P. Palsson et al. (1993). And further, we investigate properties of GMVC for time-varying system. In order to avoid consuming the vast energy, in general, the control-weighting of cost function is given as a positive value. But this may cause the offset in the output of the system. Also, the offset may occur because of input disturbance. To eliminate these offsets, in this research, the technique that establishes a compensator based on the internal model principle is proposed, that is the servo system materialized against the reference signal change and load disturbance.
A fuzzy based circuit to recognize circular patterns is proposed in this paper. The proposed algorithm calculates the center coordinates of the target patterns. The simple algorithm and exemption from the use of template patterns as well as multipliers enable the proposed circuit to implement on the hardware of an economical scale. Furthermore, the circuit design by using current-mode techniques provides easy extendability of the circuit and efficient pattern recognition with high-speed. The validity of the proposed algorithm and the circuit design is confirmed by computer simulations. The proposed pattern recognition circuit is integrable by a standard CMOS technology.
This paper describes a case-based reasoning system using case evaluation for diagnosis tasks(called CBED). The knowledge is based on a cost expectation value which is calculated from objective and subjective values. We have been developed the quality management system, which employs a stochastic method. However, in some cases, this stochastic-based system doesn't select good cases. Therefore, we integrated some expectation values into the case selection mechanism. The CBED has some expectation measurement. Its case selection criteria use not only similarity, but also some expected values. Some unforeseen malfunctions may occur due to unsuitable design, manufacturing condition and unsuitable usage. The similarity is not enough to select useful cases from a case base. Because the similarity is based on the product models. The CBED adopts the cost effective expectation value to pick up cases. The CBED selects important repair cases based on the quality of cases, which considers repair time, repair part cost, trouble recurrence, the accuracy of reasoned cause and maintenance difficulty. We validated this system in real product repair problems.
A transportation development in recent years is quite remarkable. However, poor visibility often cause an accident. Therefore, it is very important to forecast a fog occurrence. In this paper, we propose a scheme to forecast a fog occurrence by using the Learning Vector Quantization (LVQ) and a Genetic Algorithm (GA). This scheme forecasts the fog occurrence by the weather data which are provided from the Japan Meteorological Agency. First, the provided data formation are shown. Next, the prediction scheme is described in detail. In this method, input attributes for a LVQ network are selected by real-coded GA to improve forecast accuracy. Furthermore, a partial selection processing in the real-coded GA impreoves its convergence properties. Finally, in order to show the effectiveness of the proposed prediction scheme, computer simulations are performed.
This paper focuses on the fusion algorithms of multi-layer information, in which the environment quantities of temperature, humidity and brightness are to be sequentially aggregated. The fusion result is to give an overall illustration about the environment as comfort, which has been defined as one kind of high-level information. Three multi-layer fusion methods are totally proposed in this paper; i) probability criterion, ii) rule based fuzzy strategy and iii) multi-layer creditability tactics. Different from the popular fusion strategies, the presented approach works in a step-by-step framework, and proves to be more practical and more effective especially when more variables are in calculation. The applications to a multi-functional sensor have proved that the proposed algorithms deserve the advantages of multi-layer sensing. In the light of applications, these methods are compared each other in terms of efficiency and limitations. The extension of the fusion mechanisms to multi-sensors is also briefly discussed.
It is reported that correlation of inputs in a time domain plays an important role in plasticity of synaptic transmission. Here, we studied how correlated external stimulation modified responses of the cortical networks. Rat neocortical neurons were cultured on electrode-array substrates and activity evoked by electrical stimulation from a single site was compared between before and after double site stimulation. Double site stimulation was applied either simultaneously or with time difference of 500 ms. Either increase or decrease in evoked responses was observed in seven cases of nine samples. The next step will be to understand the correlation between these changes in activity and the applied stimuli.
This paper describes a new approach for competitive learning network and its application to fault diagnosis of electric parts using image analysis. Conventional competitive learning such as RCE network proposed by Cooper et. al. has a shortcoming with its decision making process. Thus, we introduce the confidential voting process in order to avoid the ambiguous decision making in the network. Generally speaking, reliability of electric product is evaluated with some average properties such as MTTF. However, if we can find an efficient method for fault diagnosis of individual electric product, the more efficient maintenance and quality control can be performed. The basic concept of this study is that the performance degradation of carbon film resistor can be estimated from their images. We first discuss some essential issues to be considered in reliability problems. Experimental results by using the proposed competitive network with confidential voting are also described. It is shown that the new competitive learning approach give a good performance.
In this paper, a discrete-time CNN using 1-dimensional chaos circuits with controllable nonlinear functions is proposed. The proposed CNN consists of p×q 1-dimensional chaos circuits which are called cell circuits. The nonlinear functions of the cell circuits can be controlled by employing fuzzy scherne. Thanks to the controllability of the nonlinear functions, the proposed circuit can adjust transition behavior of the CNN, electronically. Furthermore, the chaotic behavior of the cell circuit which is a portion of the proposed CNN is simple since the cell circuit is a 1-dimensional chaos circuit. To confirrn the validity of the proposed CNN, two types of nonlinear dynamical phenomena are observed by SPICE simulations: chaos synchronization phenomenon and traveling wave phenomenon. The proposed CNN is integrable by a standard BiCMOS technology.
Dynamic inverse optimization generates a quadratic criterion function under which given input and output sequences become optimal for a known state equation model. In this paper we propose to use statistical test to decide which state variables are included in a criterion function. To evaluate the effectiveness of the proposed method, simulation experiments are carried out. Statistical test successfully determines which state variables are included in a criterion function.
An optical powering system has been constructed, which shows high efficiency compared to conventional systems, almost independently of the power consumption of the load equipment. This is achieved by using feedback to control the optical power supply to meet the load requirements when the power consumption changes. The maximum system efficiency is 9% and the system can deliver 300mW of electricity to a device at 200m. The system mainly consists of a laser diode, optical fiber, photovoltaic cell, and electric double-layer capacitor, as an optical power source, transmission cable, optical-electrical power converter, and electrical energy storage device, respectively. This system is particularly useful for driving remote electrical equipment that requires complete electrical isolation and/or immunity from electromagnetic noise and also has variable power consumption.
An efficient two-dimensional photon counting procedure is described for high-S/N measurements of weak spectra using an image-intensified CCD camera. By using a procedure which utilizes the signal intensities of more than one pixel in judging a single photon detection, the photon detection efficeincy was increased by _??_40% in comparison with the method where a photon detection is based on the indensity of one pixel.
Recently, the theoretical and experimental researches on the design method of optimal servo system have been done. However, the weighting matrix of cost function is decided on desired value by trial-and-error. Therefore, it seems that a systematic method to decide weighting matrix is necessary. In this paper, one method which gives the constraint in upper bound of the tracking error and decides the weighting matrix satisfying the constraint is proposed. Though the algorithm for decision has been used for regulator problem until now, in this paper, we use it in servo problem.
We theoretically investigated the optical forces exerted on a sphere in order to corroborate the threedimensional optical trapping of a micro object by the laser heaths from plural optical fiber ends. which were inserted at an angle to a sample chamber. From these investigations, we verified that there was only one stable point of equilibrium located below the bean-crossing point and greatly enhanced stability could be obtained.