The impact of the convected charges, which is caused by the electrification phenomina, in the oil-paper insulation systems of power transformers has been numerically investigated using FEM. The potential distribution of the oil-paper insulation system is assumed to be governed by three equations instead of the classical Laplace equation. When the charge density of oil and the conduction current are kept high, the field distortion or the interrace in a power transformer is considerably distorted due to the convected charges.

In this paper the instability analysis of Inverter fed Induction motor drive system is performed by calculate the eigen value of the linearised equation which describe the behavior of the Induction motor. Inverter fed Induction motor drive system may become unstable at low speeds(frequencies) even though balanced, sinusoidal voltage are applied. Effect for the change of machine parameter and Inverter delay time are simulated by digital computer.

The Complex Harmonic Balance Finite Element Method CHBFEM ) is dicussed for the time - periodic magnetic field with saturation characteristics. And Jw - method which is used for analyzing liner system with sinusoidal voltage input can be generalized in nonlinear time-periodic magnetio field system. The CHBFEM enables us to calculate the each harmonic magnetic flux ditribution and the distortion of currents resulting from material at an AC voltage source and to save calculating time, the number of calculation and computer memory.

In this paper, perpendicular magnetic recording system which consists of a single pole head and a double, layer medium is analyzed by using 2D-finite element simulation and Preisach model for describing hysteresis characteristics. For confirming this method, we compare the calculated magnetic field intensity which is above the isolated transition of magnetization and scale up experimental data which have already published. The results show a good agreement and we can conclude that this method is very useful to analyze perpendicular magnetic recording system.

In order to optimize systematically the shape of electromagnetic devices, two sensitivity analyses, one based on finite element method and the other based on variational formulae and material derivative in continuum mechanics, are proposed. These are applied to eddy current problem of time-varying field and verified these usefulness.

This paper proposes the calculation method of the needed thrust force and the input current which are considered the end effect and the normal force effect of SLIM for the propulsion vehicles. To estaninate theperformance of SLIM, the thrust force, the normal force, the efficiency and the power factor are comput ed according to the variation of frequency and pole pitch.

Electromagnetic levitation system became power-full because of Its high speed, safety, reliability, and so forth So, in this paper, various characteristics of magnet in electromagnetic levitation system were analyzed by Finite Element Method and its design parameters were proposed by simulation results.

If one express the basio equations of electromagne tics in terms of differential form, one can have many physical meanings. To obtain this advantages in Finite Element Method, we should use new element. In this study, we select degree of freedom as circulation A along edges of the mesh, and use Egde Element because A is i-form. We apply this method to some examples of the 3-D magnetostatics, and obtain decrease of total nonzeros and increase of accuracy.

According to the rapid growth of transmission power, fault current level approaches to the maximum capacity of circuit breaker. This paper has been proposed a superconducting current limiter (SCL) to reduce the fault current to appropriate level. Superconductor for SCL has been examined and design criterion of each triggering method has been shown. Superconducting current limiter was fabricated using thin film and switching characteristics were shown, where thermal and current triggering were adopted.

A general scheme for a discrete adaptive observer having exponetial weighting properties is presented for a single-input single-output linear system. In this scheme, all the past measurement data are weighted esponetially both with the weighting factor and the stable matrix F. This observer is then implemented in the design of an indirect adaptive pole placement contoller. To increase nemerical stability in getting the controller parameter, a recusive algorithm is introduced. It is shown that the overall control scheme is globally stable with the persistent excition

Deafs cannot speak straight voice as normal people in lack of feedback of their pronunciation, therefore speech training is required. In this study, fundamental frequency, intensity, formant frequencies, vocal tract graphic and vocal tract area function, extracted from speech signal, are used as feature parameter. AR model, whose coefficients are extracted using inverse filtering. is used as speech generation model. In connect ion between vocal tract graphic and speech parameter, articulation distances and articulation distance functions in selected 15-intervals are determined by extracted vocal tract areas and formant frequencies.

A parametrization for a linear system is presented to design a direct model reference adaptive pole placement controler. This parametrized model is one of the structured nonminimal models. The exponentially weighted least-squres algorithm is employed to estimate the control parameters. The direct adaptive controller has the exponential weighting properties by the proposed method of selecting the characteristic polynomials of the sensitivity function filters in connection with the reference models.

This study presents new algorithm for generating octree by volume intersection with pseudo-octrees which are constructed for each orthogonal face views in order to represent 3D objects exactly. In this algorithm, optimal quadtree is constructed for depth images in which distance information is represented in gray level, and then pseudo-octrees are constructed by using optimal quadtree in each orthogonal face view. It is shown that 3D object with concave parts can be reconstructed exactly by our suggested algorithm, and that memory space and tree construction time are also reduced compared with other algorithms.

This note presents a direct adaptive pole assignment control for general discrete, linear, time-invariant, nonmimum phase system.Controller parameters are estimated from the recursive least-squares algorithm, and some additional auxiliary parameters are obtained from aset of recursive equations based on a certain polynomial identity which is derived from the pole assignment equation and the Bezout identity. This scheme increase the numerical stability of the auxiliary parameters, and guarantees local convergence without any extra conditions for the external input. The effectiveness of the proposed scheme is demonstrated by the computer simulation.

A parallel processing algorithm for discrete Kalman filter, which is one of the most commonly used filtering technique in modern control, signal processing, and communication. is proposed. Previously proposed parallel algorithms to decrease the number of computations needed in the Kalman filter are the hierachical structures by distributed processing of measurements, or the systolic structures to disperse the computational burden. In this paper, a new parallel Kalman filter employing a structure similar to recursive doubling is proposed. Estimated values of state variables by the new algorithm converge with two times faster data processing speed than that of the conventional Kalman filter. Moreover it maintains the optimality of the conventional Kalman filter.

In this paper, it is introduced about Auto-Tuning Method which applied to Expert Controller. This method constructs heuristic knowledge of the tuning experiment in order to realize relay tuning theory of Astrom &. Hagglund. This method applies to various Plant and showes utility which is used for Controller design of PID group. Result of simulation showed superiority for speed reponse and robustness of all Plant.

A new statistical linearization technique for nonlinear system called covariance matching method is proposed in this paper. The covariance matching method makes the mean and variance of an approximated output be identical real functional output, and the distribution of the approximated output have identical shape with a given random input. Also, the covariance matching method can be easily implemented for statistical analysis of nonlinear systems with a combination of linear system covariance analysis.

This paper presents a continuous-time robust direct adaptive algorithm in the presence of bounded disturbances and / or unmodeled dynamics. In the new algorithm, Narendra's adaptation law is adapted. And a term, proportional to the product of tracking error and normalizing signal, is added to the conventional control law. It is shown that the performance of the adaptive schemes is improved if a proportional adaptation tera is added to the control law. The scalar case is only discussed in the stability analysis. Computer simulation is presented to complement the theoretical result.

This paper presents a method of sub-optimal control for nonlinear systems via block pulse transformation. The adaptive optimal control scheme proposed by J.P. Matuszewski is introduced to minimize the performance index. The proposed method is simple and computationally advantageous. Viablity of the this method is established with simulation results for the van der Pole equation for comparision with other methods.

Regional electricity demand forecasting is among the most important step for lone-term investment and power supply planning. This study presents a regional electricity forecasting model for Korean power system. The model consists of three submodels, regional economy, regional electricity energy demand, and regional peak load submodels. A case study is presented.

It is important to analize voltage drop exactly in distribution systems. However the average length of a feeder is about to 40 km long and the exact modelling of a sectional share is very difficult. An efficient simplified model is necessary for dealing with such a long and complicated feeders. Especially, distribution feeders are linked to customers directly and maintaining the end voltages within the regulation is very important. This paper introduces distribution load factor for simplifing the complicated feeders in a proper manner. Test results show the more enhancement of accuracy and the better applicability in field sense.

Recently, electric energy consumption pattern shows very high peak load with low load factor. This Load pattern have made electric utilities be interested in Load Management, many studies are reported. But most of these studies are concerned with Rate - Load shape relation, a few of these are concerned with generating cost reduction. In this study, the effect of Load Management is incorporated to establish optimal Generation Expansion Planning. Using avoided cost, optimal generation expansion planning is achieved to make maximum avoided cost of Electric utility. Dynamic programming technique is used to solve this algorithm.

Associated with the protection and control functions in the substations, a new network scheme based on the IEEE 802.4 is proposed in this paper. This network is designed for supporting 3 data classes: 1st, it includes the highly time-critical data for the fault detection function. 2nd, routine data with a period is to served for monitoring the power status. Finally, it contains the control data which is used for reporting an event about protection or tuning some values related with power status. The way to allocate target_token_rotation timer is provided to guarantee the protection and control activities in the suggested network for the substation.

본 논문은 디지탈 거리 계전 방식의 구현을 위해 최적 FIR 필터를 사용하는 알고리즘을 제시한다. 이 알고리즘은 고장 발생시 과도 현상에 의한 임피이던스 궤적의 영향 면에서 볼 때, 기존의 디지탈 거리계전 알고리즘이나 Kalman 필터를 이용한 거리계전 알고리즘 보다 훨씬 유리하여 고장 발생 시 빠르고 정확한 고장 판단과 고장점의 산출을 제공한다. 이러한 알고리즘은 EMTP를 통해 생성된 고장 데이타의 실시간 적용을 통해 입증되었고, 이의 실시간 구현은 모토롤라 DSP (Digital Signal Processor) 56001을 연산 전용의 프로세서로 사용한 디지탈 보호계전 시스템을 통하여 성공적으로 수행되었다.

This paper presents a new aggregation method of induction motors for the load flow and transient stability analysis. The proposed method aggregates a group of motors into a single motor considering the variations of the system voltage and frequency so that the aggregated model represents the dynamic characteristics of the system accurately. The performance of the Method is shown by numerical simulation and is compared with that of LOADSYN algorithm.

Distance relaying is one of the most commonly used protection schemes in the high voltage transmission systems. In this scheme, coordination between the primary and backup relays is very critical for the power system security. This paper reports an rule-based methodology for checking and correction of miscoordination problems in the distance relaying. Diagnosis rules achieving an accurate but simple checking have been developed through the geometric analysis of the impedance characteristics of the distance relays. Heuristic rules having the pratical power for miscoordination correction are suggested. The proposed method has proved very effective through the several case studies on the actual systems.

A program for the protection coordination of distribution systems is reported. The developed program nailed DISCO, consisting of two modules has a capability of selling the overcurrent relays, reclosers, sectionalizers, fuses which are commonly used in the primary feeder systems. Setting is performed in a downstream fashion starting from the substation, and coordination rules are applied identifying the pattern of the primary-backup devices based on the depth-first search. T/C curves of various devices are handled by digitization for accuracy and various graphics and windows are adopted for user-friendly interface.

This paper reports an expert system for coordination of protective relays in the high voltage transmission system. The proposed system consists of five modules and has adopted the frame and production rule representations achieving the efficient data storage and knowledgebases. It has an interface to the fault program PSS/E and has reduced the data retrival time by implementing the local database containg only the minimum information for the process. Different relay parameters and output formats of the setting results can be easily incorporated owing to the rule-based structure. Graphics-based output helps understanding of the process and enhances the practical power.

For transient stability analysis of a power system, the new method using transition matrix is introduced in this paper. At the present the, Runge-Kutta, Modified-Euler and Trapezoidal methods have been very popular in most stability programs, Modified-Euler and Trapezoidal methods are inferior in accuracy and Runge-Kutta method has problems in computation time. The proposed algorithm requires transition matrix and its integrated values with derivatives of nonlinear parts in nonlinear differential equations for stability analysis. The method presented in this paper is between Modified-Euler and Runge-Kutta methods from the view point of computation time and is superior to the other methods in accuracy.

This paper presents a new approach to optimal power flow (OPF) problem using conjugate gradient method, using this method. We can obtain initial feasible solution and lagrangian multiplier without calculation of matrix inversion.Test experiment shows a desiriable result and a stable convergence characteristic.

The human brain has the most powerful capabilities in thinking, interpreting, remembering, and problem-solving. Artificial neural network is appeared by scientists who have tried to simulate such a human brain. The artificial neural network has the capability of learning, massive parallelism capability and robustness for disturbance which are necessary for power system application. In this paper, We reviewed the typical topologies and learning algorithms of artifical neural networks which can be used for pattern classification. And we surveyed for the applications of artifical neural network to the power system.

This paper treats modeling of dynamic load characteristics for power for systems. The dynamic load is represented as 4th order multivariable ARMA model under the assumption that the dynamic load characteristics can be described by the dynamics of only one induction motor. The parameters of the proposed ARMA model are identified using the well-known RLS method. This paper presents two kind of identification results : one is for induction motors and the other is for field data at Donghae station. From these results, the proposed model is quite suitable for the dynamic load characteristics. It has, however, a disadvantage in the viewpoint that the identified parameters are not those of the induction motor.

This paper describes a new method of calculating expected energy generation and loss of load probability (L.O.L.P) for electric power system operation and expansion planning. The method represents an equivalent load duration curve (E.L.D.C) as a mixture of cumulants approximation (M.O.C.A), which is the general case of mixture of normals approximation (M.O.N.A). By regarding a load distribution as many normal distributions-rather than one normal distribution-and representing each of them in terms of Gram-Charller expansion, we could improve the accuracy of results. We developed an algorithm which automatically determines the number of distribution and demarcation points. In modelling of a supply system, we made subsets of generators according to the number of generator outage: since the calculation of each subset's moment needs to be processed rapidly, we futher developed specific recursive formulae. The method is applied to the test systems and the results are compared with those of cumulant, M.O.N.A and Booth-Baleriaux method. It is verified that the M.O.C.A method is faster and more accurate than any other methods.

Recently, the phenomena of voltage instability have become major concern in power system. These phenomena are closely related to what are called multiple load flow solutions and calculation methods on these solutions have developed. But conventional methods require much run time. In this paper, by using sign of |J| and weighting factor considering system configuration, fast calculation method on the multiple load flow solutions is presented.

This paper proposes a new method of the state reduction in dynamic equations of generators in large electric power system stability analysis. This method assumes study groups whose state trajectories we are interested in, coherency groups whose state trajectories are similar to the other state trajectories of generators in the same coherency group by a certain disturbance. By the weighted sum or the other method, the states of generators in one coherency group can be reduced to the equivalent states of an equivalent generator. This method is shown to be highly efficient in reducing the number of states with small error by the result of case study presented latter part of this paper.

This paper presents a preliminary study of introduction of the Dempster-Shafer inexact reasoning method to the expert system for the power system design problem. A brief review of Dempster-Shafer theory of evidence is presented and development of an inference engine adopting the Dempster-Sharer theory is reported. Developed inference engine has a ability of handling both the confirming and disconfirming knowledge represented in the production rule, and has a general purpose application in the design and diagnosis problems. Its applicability has been tested on the problem of the protection scheme selection, one of the typical design problem and we believe, it has shown the feasibility of adoption of the inexact reasoning methodology into the design problem.

When load areas on a feeder are deenergized due to faults,operaters need to identify nelghboring feeders, try to restore customers and minimize out-of-service areas. These cases include knowledge of system states and various constraints such as voltage drop. This paper concerns the load transfer infault restoration. Also, it is considered the operating constraints such as line current capacity, relay trip current, transformer capacity, voltage drop and line loss. The expert system is able to propose the optimal load transfer method by anallzing system states and considering constraints.

Recently, it is reported that the behaviour of PECVD under high electric field and current condition has a major effect on MNS device degradation. In this paper, we evaluated the breakdown and TDDB characteristics of Laser assisted PECVD SiN films which is introduced new deposited method. And also, long term insulator breakdown reliability is described by examing time dependent dielectric breakdown under positive voltage. Failure tines against electric field are examined and acceleration factors are obtained for each case. From these data, breakdown wearout limitation for Laser Assisted PECVD SiN film can be characterized.

Today, according to the temperature lowering of VLSI technology which have been required, the new thin film technology of low temperature have appeared. Plasma CVD method, one of low temperature technologies, have major problems with many interface trap defects. In this paper, we prepared ammonia free SiN film containing small H that acts as a defect impurity, and investigated the electrical properties of Laser assisted deposition film.

The lattice mismatch between ZnS and Si is negligible because of its value being 0.39%. In this study, $ZnS/_{(P)}Si$ solar cell were fabricated as a layer of ZnS is epitaxially grown on a silicon substrate by PVD method and its photovoltaic properties were measured and discussed. The heat treatment was done after deposition. As the temperature increased up to a certain value, the film has better perfection in crystal structure and electrical characteristics. Measurments of the change that occur in the ZnS films were made by SEM., X-ray diffraction. The optimal thickness of film showned $0.6{\mu}m$, being measured by SEM. The great improvement of the grain growth ZnS film came out after heat-treatment. The result obtained from the $ZnS/_{(P)}Si$ solar cell as follows:short circuit current; $I_{sc}=54mA/cm^2$,open voltage; $V_{oc}=400mV$, fill factor FF=0.72, conversion efficiency; ${\eta}=15.6%$ under the irradiation of 100 ($mW/cm^2$) focused by solar energy. And these are discussed in comparison with other kinds.

A new technique is described for developing low-cost $SnO_2$ : F thin films as TCO (Transparent Conducting Oxide) substrate of a-Si solar cells. A novel Pyrosol equipment has been developed, and $SnO_2$ : F thin films have been deposited under the condition of varing dopant concentration, temperature and composition rate of solution. Futhermore, electrical and optical properties of thin films have been measured, and exhibit resistivity of $4.3{\times}10^{-4}{\Omega}$ cm and transmittance of 80% which is almost at the same level as those of $SnO_2$ : F thin films by CVD.

A $SnO_2$: F/$SiO_2$ duble layer on the sodalime glass is described for developing a low-cost substrate of a-Si solar cells. Dipping and Pyrosol method, have been used for thin film deposition, and electrical and optical properties have been analysed. Finally, p-i-n a-Si solar cells have been fabricated on this substrate by plasma CVD and their average efficiency is 4% approximately.

In order to investigate the radiation effects induced to electrical properties of Biaxially-Oriented Polypropylen film, several observations were carried out to the sample irradiated to various dose by $Co^{60}-{\gamma}$ ray, on the characteristics of TSC spectra measuered as a function of electric field applied to a sample of 15[ ${\mu}m$] thick. The TSC spectra observed in the temperature range of $153{\sim}403[K]$ with the electric field of intensity $10{\sim}60$ [MV/m], have shown two of the distinguished peak such as ${\beta}$, ${\alpha}$, each of which appeared at $-5{\sim}20$ [ $^{\circ}C$] and 90 [ $^{\circ}C$] respectively. As the conclusions, obtained from the studies, the origin of ${\alpha}$ peak in TSC seems to be attributed by thermal excitation of ions trapped with $0.4{\sim}0.8[eV]$ deep, at the defects formed by $Co^{60}-{\gamma}$ irradiation in a crystaline region. The origin of ${\beta}$ peak was regarded as the depolarization process of "OH" or "CO" dipole with the activation energy of $0.4{\sim}0.6[eV]$ in an amorphous region.

The trap depth and its density of highly elongated polyethylene have been studied by thermally stimulated current measurements. Three discrete traps with depth of 0.13 0.3 and 0.5 eV have been evaiuated using the initial rise method in the temperature region 77 to 390 K ; these change slightly with the elongation ratio. A trap density of the order of $10^{13}-10^{14}/cm^3$ has been calculated from the area of TSC peaks.

The existing batteries such as lead acid battery and Nicad battery have been at serious issue, because heavy metal such as Pb and Cd give rise to environmental pollution. Therefor, when these material is changed over polymeric electroactive material, we expect environmental pollution will be prevented. We decided to develop 2025 coin type PAn/Li-Al secondary battery for goal of memory back-up battery. This report is concerned with manufacturing of 2025 coin type PAn/Li -Al secondary battery.

The purpose of this study is to research and develop polymer secondary battery. This paper describes the first discharge characteristics of PAn/Li-Al secondary battery. PAn was prepared in $HBF_4$ aqueous solution by galvanostatic electropolymerization and then used as cathode active material. PAn/Li-Al secondary battery was prepared in 2025 coin type. Characteristics of this battery are summarized as follows. ${\bullet}$ Open curcuit voltage and discharge end voltage was 3.5V and 2.9V, respectively. ${\bullet}$ The ratio of electricities in discharge to theoretical electricities in all undoping of PAn cathode was 56% at constant current discharge of 1mA. ${\bullet}$ The capacity density, energy density and maximum power density per weight of PAn electroactive material were 56.1Ah/kg, 168.4Wh/kg and 16.9kW/kg, respectively.

The cordierite(2 MgO, $2Al_2O_3$, $5SiO_2$) is of great interest for microelectronic substrate of multilayer intergrated circuits. The metal used in this study was copper, and metal layer is fabricatedon the cordierite substrate by the screen printing method. We studied the adhesion properties of the interfaces due to the different cosintering conditions. When cosintering in the $Ar+H_2O$ gas, the adhesion is very good. Specially heating rate is very important factor for the adhesion.

A.C. impedance properties of $RuO_2$ based thick film resistors which having different resistivity value (DuPont 1721 : $100{\Omega}$/ sq., 1741 : $10K{\Omega}$/sq.) were investigated using by impedance analyzer. In case of lower resistivity 1721 system, the complex impedance was composed nearly R component for all speciman sintered at above $600^{\circ}C$, and the frequency dependancy on impedance was not affected very much up to 5MHz and again gradually increase with increasing the frequency. In case of higher resistivity 1741 resistor system, impedance properties were very depandant on sintering temperature. When sintering temperature was $600^{\circ}C$, the complex impedance plot shows a vertical line, which correspond to lone capacitance equivalant circuit, and the impedance linearly decreased with increasing frequency. In case of speciman sintered at $700-900^{\circ}C$, the complex impedance plot shows semi-circular are correspond to a lumped RC combination, and the impedance shows constant value to 5MHz, again decreased with increasing frequency. But the complex impedance behavior of speciman sintered at $1000^{\circ}C$ was shows the equivalent circuit correspont to parallel combined LCR component, and the impedance was not varied with frequency.

In this study, we have studied the optioal properties and $I_2$ doping effect of Poly(3-Akkylthiophene) solution. The end of speotrum absorbance on the solutions of Poly(3 - Octylthiophene), Poly (3-Dodecylthiophene) and Poly(3-Dooooylthiophene) was dependent on temperature. It is bellved that conrod-coil transition dependent on temperature was rod-coil transition. And absortance peak of Poly (3-Dodecylthiophene) in dichloromethane solution increased with increasing the $I_2$ dopant.

The addition of titanium has come to produce an increase in the conductivity of ${\alpha}-Fe_2O_3$ and has been shown NTC ( negative temperature coefficient ) characteristics. Titanium enters the ${\alpha}-Fe_2O_3$ lattice substitutionally as $Ti^{4+}$,thus producing an $Fe^{2+}$ and maintaining the average charge per cation at three. Thus the $Fe^{2+}$ acts as a donor center with respect to the surrounding $Fe^{3+}$ ions. The sintering temperature, compacting pressure and sintering tire have an effect on the electrical properties. C-V and other properties have been measured on polycrystalline samples of ${\alpha}-Fe_2O_3$ containing small deviations from stoichiometry and small amounts of added Titanium. This measurment was made in the course of an investigation of the NTC mechanism in oxides whose cations have a partially filled d-level. C-V and frequency properties have been applied to the measurement of the trap barrier properties at the grain boundary. The double Schottky barrier at the grain boundary is the major cause of the NTC mechanism in NTC thermistor of ${\alpha}-Fe_2O_3$ containing N-type impurity.

Soft magnetic silicon steels containing up to 6.5wt% of silicon were prepared by powder metallurgical processing and their magnetic properties were evaluated. The magnetic properties of P/M silicon steels are similarly affected by the silicon addition as those of conventional ingot processed ones but are also significantly affected by density and interstitial impurities particularly oxygen content. Magnetic flux density, $B_{10}$ and coercivity, Hc, tends to decrease with silicon content whereas maximum permeability, ${\mu}m$, decreases first and then increases rapidly above 5 wt% silicon. Increasing density also increases magnetic flux density and maximum permeability but reduces coereivity. The latter two properties are, however, affected more strongly with oxygen content.

For improving magnetic properties of the amorphous Fe-B-Si alloy, we annealed in a magnetic field oriented in the plane of the ribbon longitudinal to its long axis. By field annealing, coercive force and total core loss are reduced from 0.04 Oe to 0.02 Oe, and from 0.25 watt/kg to 0.15 watt/kg respectively comparing with non-field annealed specimen. These reductions were caused by the formation of $180^{\circ}$ domains parallel to the annealing field due to the induced anisotorpy.

The effect of $CaCo_3$, $SrCo_3$, $SiO_2$ and $H_3BO_3$ in the range of 0.1-1.0 wt% on strontium ferrites consisting of the magnetoplumbite phase $SrO-5.7Fe_2O_3$ were investigated. The hysteresis loop, density, demagnetization curve and the intrinsic coercive force were measured on anisotropic ferrite. The particle diameter and 0.5(wt%) of second additive $CaCo_3$ is particularly important for the properties of anisotropic ferrite. When the particle diameter is decreased from 1.98(${\mu}m$) to 1.07(${\mu}m$), the remance is increased from 2900 to 4010(G) and the coercive force from 2150 to 2850(Oe) at a sintering temperature 1230 ($^{\circ}C$). Remance Br(G), coercleve force(Oe) and maximum energy product of sample A-16 are87(%), 56(%), 67(%) of S-W model theoretical value.

The theory was analyzed and the computer program was developed for simulation about ceramic filter in the state of the one-strip electrode using the energy trapping effect. They were fabricated using the PZT-4 specimen. Each of the average differances of the resonant point and bandwidth between by the theoretical calculations and by experiment results was 5.6[%] and 3.72[%], and it is considered that this differance is originated from the high temperature and stress during the process that the ceramic filter was fabricated It is considered that the one-strip ceramic filter having a wanted characteristics and the lowest differance is fabricated easily by means of the simulation developed in this paper and the fabrication methods.

Nondestructive eddy current testing has been used for measuring the electromagnetic properties of metals. It is well known that the conventional eddy current sensors are not effective to measure multi-layer properties. In this paper, the mutual inductance type sensor is studied and transformed fractions of metals are measured. Characteristic factors in eddy current testing are considered and their influences on the testing results are investigated.

A single-crystalline epitaxial film of GaAs has been grown on Si using an ionized cluster beam technique. The native oxide layer on the silicon substrate was removed at $550^{\circ}C$ by use of an accelerated arsonic ion beam, instead of a high-temperature desorption. During the growth the substrate temperature was maintained at $550^{\circ}C$. Transmission electron microscopy and electron diffraction data suggest that the eats layer is an epitaxially grown single-crystalline layer.

We fabricated a high Tc superconducting motor named the "Meissner Motor" using Y system high Tc superconductors, which rotates with the repulsive force caused by the Meissner effect. And, we derived the optimal rotating conditions and parameters from the examination of its chracteristics. Finally the possibility of a high Tc superconducting rotor has been investigated.

Monte Carlo simulations of electron impact ionization in silicon utilizing simple spherical band approximation including L valley are presented. The mean energy of electron at high electric field is lowered and the threshold energy of electron impact ionizaiotn is smeared out to $E_x$ by including L valley. This work also presents the importance of the modeling of L valley by calculating the electron population rate of it over the threshold energy of eletron impact ionization, which is higher than 10%

The specific contact resistance(SCR) of metal-semiconductor interface is an important design parameter for VLSI interconnecting technology. As the critical feature size of the integrated structures decrease, the physical size of ohmic contacts will also decrease and the series contact resistance will increase. Al-Si contacts on the annealing condition are studied. The propreties of the contacts depend considerably on the annealing procedures. Barrier height is measured from Capacitance-Voltage characteristics. The specific contact resistance are analyzed using a modified four point method.

In this study, PZT powder prepared by the wet direct process was synthesized. As starting materials, $TiCl_4$. $ZrOCl_2{\cdot}8H_2O$ and PbO were used. Uniformly shape and fine-grained PZT powder was obtained by the wet-direct process and PZT powder was characterized by XRD, DTA analysis. The X-ray diffraotion peaks from the PZT powder were observed at 700($^{\circ}C$) or over.

This paper introduces a technique of shifting the resonant frequency and of depressing the spurious modes in the composite piezoelectric vibrator laminated in the order of piezoelectric vibrator/thin insulator/piezoelectric vibrator, whose vibrators have the the same sizes and the same properties, respectively. The composite vibrator has shown the characteristics that the fundermental resonant frequency has decreased and spurious modes have depressed with increase of the capacitance of the inserted insulator. And the frequency characteristics of the composite vibrator without insulator have been nearly the same as those of each vibrator.

This study investigated electrical characteristics of solid and liquid epoxy rosins by measuring dielectric breakdown and dielectric loss when epoxy resins were exposed to a mixing cure, i.e., Fusion Blend Method. It was found that mixing epoxy resins were superior to dielectric breakdown and has shorter curing time compare with those of pure liquid epoxy resins.

This paper describes an nstrument for the detection and geometric location of partial discharge(PD) sources in transformers. This instrument measures electric current pulses and ultrasonic pulses simultaneously, counts the number of electrical pulse and determines the geometric location of PD in transformers. It was found that there is a relationship between partial discharge magnitude and pulse repetition rate when the applied test voltage and oil temperature were varied. Through the laboratory test using model transformer, it was clarified that this detector could be used, satisfactory for detecting and locating of PD in the transformer.

This paper describes on partial discharge in polyethyelen that are important for materials analysis, and so to will introduce to system of partial discharge analysis. The results of this study will be provided on properties of partial discharge in polymer materials.

This paper reports the characteristics of sparkover discharge in flowing air ranging from 0[m/s] to 30 [m/s] under the needle-needle gap. Flowing air duct of this investigation is circular tube. The important results obtained form this study are as follows. 1. the ratio of sparkover voltage to the Reynolds number decreases with increasing the Reynolds number. 2. The duration time of sparkover(t) decreases with increasing the Reynolds number. 3. the empirical equation obtained form this experiment is [ %]${\frac{Vs}{Re}}$ = A + $B{\varepsilon}^{C.Re}$ where A = 10.2 b = 125 c = -4.66 ${\times}$ $10^{-5}$

This study is to develope a spacer having almost uniform field distribution along the profile by controlling the high electric field in tensification at earthed part, especially triple junction in the spacer-electrode-gas boundary interface. Based on the extensive field calculation using CSM, a model spacer has been manufactured and tested. The test results show that the breakdown strength of the spacer is almost same as that of SF6 gas itself without spacer.

In this work, we reviewed AC break-down voltage characteristics in $SF_6$ Gas with spacer under nonuniform field, using the electrode construction of model spacer and simple protrusion. We cleared that breakdown voltage and surface flashover strongly depend on the position of spacer's insert.

The measurement of very fast transients generated by disconnecting switches in gas-insulated switchgear(GIS) have to deal with the problems such as reliability, interference pich-up, optimal design of high voitage equipments. This paper presents a new developed voltage sensor to measure the very fast transients, the basic theory of the measuring method, the design, structure of planar capacitive voltage probe are described. Finally the examples of modelling tests on an actual site GIS are discussed.

This paper presents the test results of a model vacuum trode for the development of high power triode. The construction and test performance of the tube are included.

Referring to the calculation of the continuous permissible current rating in 22.9kV CN/CV underground distribution power cables, the current and temperature have been fully discussed and analyzed based on their three different values: one is the actually measured value throughout our test and the other two are calculated in connection with different specifications IEC-287 and JCS-168-D respectively. For this purpose, our test has been carried out with real cables which have been under stress either with induced current or with rating voltage. In the former, the calculated current of IEC-287 shows closer value to the measured one than that of JCS-168-D does. In the latter, there has been little difference on the temperature comparing with that measured without voltage application. Therefore, we think that it is preferable to choose the IEC-287 specification for the calculation of the continuous permissible current rating in the commercially power cables.

This paper presents the insulation design of 345kV OF cable normal joint which is essential to construction of underground transmission system. The insulation designs of stress relief cone, pencilling and sleeve were proposed in this paper, and the final insulation design was decided by the results of the electric field and thermal analysis using FEM. It is shown that the designed normal using proves the effect when the developing test is carried out to construct 345kV OF cable if the domestic underground transmission system.

This report the loading teat results of OF cable core in LN2 testing chamber for the development of a cryogenic cable

VSS controller is suited for systems where the robust performances in the presence of parameter variations and disturbances are most important. The practical realization of such robust VSS algorithm using DSP ($TMS_{32020}$) is dicussed in this paper. PWH-like control signal which is produced by directly using VSS control algorithm substitutes the PWM amplifier. And the armature current is used in construction of switching function for the purpose of increasing switching frequency. In order to compensate steady state error produced by various external disturbances practical integral compensator is considered.

In this paper describes the feasibility study on a single-phase / three-phase voltage-fed inverter using 1GBTs, incorporating voltage-mode quasi-resonant DC Link and the instantaneous Sinusoidal Current tracking adaptive control implementation under zero voltage switching resfriction. The simulation and experimental results are shown here and discussed for single-phase voltage-fed and three-phase voltage-fed inverter in order to verify the new operating principle of the Inverter topology proposed here.

A new quasi-resonant do link inverter is suggested, which can operate at the constant peak do link voltage irrespective of the magnitude of load current. The inverter is analyzed by using the topological analogy between the proposed inverter and the resonant DC/DC converter. Based on the analysis, an appropriate current controller is developed, which results in low current stress to the resonant capacitor and also enjoys the inherent capability of the current initialization of resonant inductor. For the purpose of confirming the inverter characteristics, some simulation results are presented.

A digital phase-locked loop speed control system of a self-controlled permanent magnet synchronous mortar fed by a voltage source inverter is presented. This paper discribes the hardware and software design of the system. Variable speed control system for self-controlled permanent magnet synchronous motor is proposed. Simulation results demonstrate the validity of proposed methods. This proposed control technique is implemented by using a microprocessor-based system.

This paper discusses the application of VSCS(Variable Structure Control System) to position control of a trapezoidal type brushless DC motor. In order to simplify the overall control system and to improve the robustness, a new switching function which is composed of linear combination of only measurable state variables Sr(x) and Sr(x) is defined. The proposed new switching function is implemented using a digital signal processor(DSP). A general PWM amplifier is replaced by an ON-OFF pattern generator for the hardware simplification and digitalization. Experimental results are given to demonstrate the validity of the proposed control method.

This paper presents the three section sliding mode control algorithm based on hysteresis current control add indirect field oriented control method, and applies it to the position control of induction motor. The three section sliding trajectories are defined in such a way that the system responds following a max acceleration line, then a max speed line, and finally a max deceleration line. This control scheme solves the problem of robustness loss during the reaching phase that occurs in conventional VSC strategy, and ensures the stable sliding mode and robustness enhancement throughout an entire response. Also, the PID controller operating in parallel is adopted to eliminate the sliding mode's collapse phenomenon near the origin caused by steady state chattering phenomenon Digital simulation results confirm that the dynamic performance of the system is insensitive to parameter variations and disturbances.

This paper proposes a novel current control strategy for current regulated VSI-PWM transistor inverter. The conventional hysteresis control method has good dynamic responses. But the switching frequencies are high because it does not optimize switching patterns. Proposed current control strategy can optimize switching patterns. As regulator, three level comparators are used. The outputs of the comparators select appropriate inverter output voltage vectors via switching table stored in EPROM. The simulation and exparimental results in comparison to contentional hysteresis strategy are presented and discussed.

In this paper, we proposed a software strategy of PWM inverter, and could indentify its realization by simulation and experiment. A basic concept of software is that the PWM pattern of inverter decided by the flux vector of induction motor tracing circle locus. The algorithm is charterized as a simplification of control circuit, abilitry on expressing control quintity as quantization and conpensation of quantization error and a switching pattern is decided by addition and substraction and high performance control is realized by 8 bits microprocessor.

This paper is an application of Modern control Theory to the control of a current-source inverter (CSI) fed Induction Motor Drive System. A Linear Quadratic-Gaussian (LQG) scheme is developed, in which the Kalman Filter is tuned for high robustness by a method due to Doyle and Stein(8). The design is carried out for a sample system and the robustness by analysis and computer simulation.

In the indirect vector control method, the rotor resistance variation caused by operating temperature change is an undesirable nature. A new adaptation algorithm to compensate for the rotor resistance change based on the on-line estimation of field vector which requires the measurements of stator voltage and rotor speed is presented in this paper. Also minimum variance controller is presented for the adaptive control performance. This algorithm has been tested by simulating the induction machine using a digital computer and the results are discussed.

The photovoltaic inverter is a system which transforms the solar energy to the electrical energy. Because the output of photovoltaic inverter depends on solar radiation, it is necessary to save generated power or to interface with utility line. This paper describes a 2KVA photovoltaic inverter utilizing IGBT, designed to have a very high efficiency in stand alone mode and to have the function synchronizing the phase between inverter and utility voltage as a fundamental study for utility interactive mode.

A sinusoidal PWM type inverter using 1GBT and a new instantaneous current tracking control stategy is described. Its utilization as an interface for a grid-connected photovoltaic system with maxmum power tracking capability is explaned, and competer simulation along with waveforms are discussed. Finally, experimental results are analysed by state space avarge method to actual grid-connected systems.

This parer presents a cycloconverter with a LC resonant circuit for a induction motor drive. The cycloconverter can provide variable voltage and frequency three phase output while keeping the input displacement factor at 1.0. The input current wave for is are sinusoidal,and the wide output frequency range is appropriate for variable speed AC motor drives.

This paper concerns a DC-DC converter using the characteristic of series resonance. Operation principle of the system described by the proposed equation is illustrated. Characteristics of steady state of the system which is essential to system design is evaluated for frequency and pulse width characteristic using dimensionless parameter. The proposed circuit based in constant voltage control uses the Pulse Width Modulation - Time Ratio Control method.

In this papers, SPICE which is widely used in analysis of general circuit and simulation in electric and electrical field is applied to DC-DC converter. Be selected Forward which is widely used as SMPS(Switched Mode Power Supply) amang DC-DC converter, and showed the aveformer, we know that the converter is operated in normal. Including the control circuit in converter, by controlling the duty ratio of switch, we know that the output voltages is staple from the transient state of convwrter. Also, comparing SPICE simulation with experiment, the validity of simulate is showed.

This paper represented constant voltage power source equipment using voltage fed inverter. This system drived syncronizing power source, charged battery connected to DC side, and constantly maintained regardless of power source voltage variation and load variation. Output made good wave with PWM method and we proved stabile constant voltage output when AC power source is failure.

All the existing PWM voltage-controlled inverters, regardless of the advantages of the operation, have produced serious problems such as acoustic noise. In this raper, the PMW technique for acoustic noise reduction is implemented using randomly modulated carrier modulation. By modulating the triangular carrier in sinusoidal PWM with band-limited white noise, it is shown that the acoutic noise level and spectral density is reduced in a tone. This method shows great merit in application the AC motor drive system to particularly environment.

A new control method of solar cell output using PWM chopper is described. Theoretical analysis and comparision for all sorts of system are also discussed for improving photovoltic system efficiency. It is expected that the result of this study will be utilized to some other DC or AC application equipments.

This paper is study on resonance type high-frequency inverter using self turn-off devices. The power conversion circuits adopt full-bridge of voltage-fed type. IN the circuit analysis, resistance load was used to estimate of characteristic.

Series connection of power switching semiconductor elements is unavoidable when a high voltage convertor is aimed. However, it is important to equalize distribution of turn-off voltage because the switching elements have different characteristics. In this paper optimal switching control algorithm is proposed so that series connected poker switching semiconductor elements can be always switched simultaneous turn-on and turn-off.

Various sampling methods are used for microprocessor-based measurement and analysis to nonsinusoidal waveforms caused by power electronic converters. The hamonic component generates the indicating errors at the measuring instruments. This can be solved by microprocessor-based measurement and hence the microprocessor-based measuring equipment and its algorithm are developed in this paper. As a result the suggested equipment has very good measuring performances.

New asymmetrical PWM AC Chopper is proposed, which can improve the input power factor and can reduce harmonic contents at both its input and load sides. The switching scheme for the proposed technique is presented and the theoretical characteristics of AC voltage regulator controlled by APWM AC chopper are investigated through the digital simulation

In this parer, the parameter for control variable is consist of value that take measured the input line current and input total power-factor, and the compensation susceptance is calculated by its parameters. The compensator can reduce the reactive power generation by the system to zero, thus the input total power-factor is kept on sustaining the unit power- factor.

In order to spread the practical utilization of solar energy, it depends upon how we can increase the efficiency of solar energy conversion system. This paper describes the method for calculating of optimal passive elements values in Maximum Solar Energy Tracking System. And experimental results with those calculated values are presented.

This paper concerns a method for high precision torque control of induction motor supplied with power via a Transister inverter. The control system is totaly digitized using DSP - TMS32010.Since operation time leads to control delay in such a system, torque ripples are more significient than relieved considerably by using current preview control and vector pulse width control. The usefulness of these techniques will be demonstrated by showing the results of a simulation,and results of measurement on a 2.2Kw induction motor.

An optimal Stabilization technique for a bilinear in duction model is introduced. This technique includes to o parts; the one is an stabilization control using Lyap unov Function which has the form of a sum of linear and quadratic function of the state variables, and the other is an optimal control using the performance index which depends on the choice of the elements of the Ly apunov matrices concerning both the state variables and the input variables. Therefore, induction motor is drived with the shorter transient time of the state variables and with the smaller overshoot of the ones, simulation results are obtained from a digital computer. Experimental ones are obtained from implementation of the optimizing controller using 8086 microprocessor kits and analog circuits are compared.

GaAs MESFETs with the submicron gat are fabricated. $G_{m,mas}$ = 195mS/mm with the $0.5{\mu}m$ gate length and $G_{m,mas}$ = 170mS/mm with the $0.6{\mu}m$ gate lenth. $f_{mas}$ = 7GHz with the $1.5{\mu}m$ gate length and the $120{\mu}m$ gate width. We can estimate that $f_{mas}$ = 15GHz with $0.6{\mu}m$ gate length and that $f_{mas}$ = 18 ${\sim}$ 20GHz with the $0.5{\mu}m$ gate length.

Comparison between two approaches for analysis of nonuniform transmission line with linear taper is presented. The input impedance evaluated analytically for tapered angle $2^{\circ}$ is deviated about 2.5% from that of numerical approach. The validity of our analytic method is illustrated physically.

Nonlinear protection of overcurrent induced on parallel two-wire transmission line by electromagnetic pulse is analyzed. This algorithm is implemented numerically and simulation result for simple incident pulse and nonlinear type agrees with its analytic solution.

Room temperature cw operations of a standing wave and a unidirectional Ti:Sapphire laser pumped by a $Ar^+$ laser have been demonstrated in the same pumping cavity configuration. Laser emission at 790 nm was excited with all-line $Ar^+$ laser pump for both cavities. The maximum output power was found to be 600 mW in the standing wave cavity and 210 mW in the ring cavity. Values of (83 ${\pm}$ 10) % for the internal quantum efficiency and (3.0 ${\pm}$ 0.5) % for the round-trip cavity loss for standing wave cavity are obtained from the thresholds and slope efficiencies measured with 2.4 and 11.2 % optput couplers. The measured cavity-loss value at 790 nm gives a meterial figure of merit (FOM) to be 146. Unidirectional operation of the ring Ti:Sapphire laser will be discussed.

He designed and fabricated oscillator - amplifier system respectively. Amplification system was composed of oscillator, amplifier and synchronization circuit.Synchronization was achived by optical delay and electric circuit delay. As a result, Power gain was 2- 3 with laser pulse width 5 - 6 ns.

Me performed an experiment about preionization electron number density of charge transfer type KrF excimer laser. At the total pressure of 1.8 and 2.2 atm with helium (He) buffer gas, the experimental range of the electron number density is 8-9 ${\times}$ 10 cm The distance between electrode and preionization pin is 15 mJ at charging voltage of 27kV,gas pressure of 2.2 atm,gas mixture of F /Kr/He = 0.3/396.7(%).

We report the development of a surface discharge preionized KrF discharge Laser with a new type of high-brightness Corona preionizer using a segmented columnar $BaTiO_3$ dielectric. This Laser characteristic was higher value than that obtained with the conventional corona preionizer using a glass, $Al_2O_3ceramic$, Kapton film, or a Teflon insulator.

This thesis presents a simple heuristic algorithm which can be applied. to a quadruped wafting vehicle for increasing the terrain adaptability. The proposed method controls a leg length which is in tranfer phase to maintain initial orientation of the robot body by using FSR type force sensors attached to foot-tips. Also, some basic experiments using the vehicle are performed to demonstrate the effectiveness of the algorithm.

In this paper, a vision-based inspection algorithm for checking mark quality on an integrated chip(IC) is proposed. In order to reduce the processing time for inspection, we are implemented image arithmetic unit and binary image projection processor in hardware. By adopting the hardwares, the processing time becomes less one sixth of that in case of using software only.

Two control algorithms for synchronizing multiple DC motors are proposed to compensate load variation and gain mismatches. Specifically, a mathematical model of a practical DC servo system is drived and analyzed to see synchronizing effect at the steady-state. Also a compensator is proposed to synchronize multiple motors at the transient-state. As an experimental system, two-axis synchronizing control system is developed and tested to show the validities of our proposed method.

A simple time-delay method for avoiding collisions between two articulated robot manipulators is proposed. Links of robot are approximated by polyhedra and the danger of collision between two robots is expressed by distances between the robots. An algorithm, which can fast obtain the minimum time-delay value needed for collision avoidance, using scheme of following the boundary contour of collision region in the collision map which has information about collisions between two robots, is described.

We propose a minimum-time path planning soheme for the robot manipulator using Hopfield neural network. The minimum-time path planning, which can allow the robot system to perform the demanded tasks with a minimum execution time, may be of consequence to improve the productivity. But most of the methods proposed till now suffers from a significant computational burden and thus limits the on-line application. One way to avoid such a difficulty is to apply the neural network technique, which can allow the parallel computation, to the minimum-time problem. This paper propose an approach for solving the minimum-time path planning by using Hopfield neural network. The effectiveness of the proposed method is demonstrarted using the PUMA 560 manipulator.

This paper introduces a path planning technique for a mobile robot in the presence of obstacles. In the technique, workspace is described by regional graph and represented obstacles by the three-layer neural network. And performance cost is defined under consideration both the traveling distance and the safety of a mobile robot. Then a collision-free path is obtained using the neural optimization technique.

In this paper, the robustness of implict self tunning controller on the continuous time system is investigated. Continuous time exponentially weighted least square algorithm is used for estimating the system parameters. The pole-zero placement method is adapted for the control algorithm. On considering the control weighting factor and realizability filter the effects of unmodeled dynamics of the plant are examined by the simulation.

Neural network is a computational model of the biological nervous system developed to exploit its intelligence and parallelism. Applying neural networks to robots creates many advantages over conventional control methods such as learning, real-time control, and continuous performance improvement through training and adaptation. In this paper, dynamic control of a six-link robot will be presented using neural networks. The neural network model used in this paper is the backpropagation network. Simulated control of the PUMA 560 arm shows that it can move at high speed as well as adapt to unforseen load changes. The results are compared with the conventional PD control scheme.

This paper presents a decentralized adaptive control scheme for multi-Joint robot manipulators based on the independent joint control scheme. The control object is to achieve accurate tracking of desired Joint trajectories. The proposed control scheme does not use the complex manipulator dynamic model, and each joint is controlled simple by a feedback controller which ensure stable and also a position-velocity-acceleration feedforward controller and also auxiliary signal, with adjustable gains. Simulation results are given for a two-link manipulator under independent control, proposed decentralized adaptive control of manipulator is feasible. In spite of a pay load variation and strong static and dynamic couplings that exist between the joints.

This paper describes the procedures for pre-tuning and re-tuning of the PI controller to specifications on patterns of output response. The key ideas of the proposed adaptive scheme are as follows. The relay feedback is adopted first for pre-tuning and the adaptive algorithms by the pattern recognition are introduced for re-tuning procedure to retune the gains whenever control conditions are changed. The proposed scheme was applied to the experimental laboratory process, heat exchanger.

In this paper, compliant motion control of a manipualator in manipulator is proposed by using the self-tuning adaptive controller. Compliant motion is needed in order to applicated to complicated and accurate fields such as assembly operation in which several parts are matched. For a control method of compliant motion hybrid control is used so forces and position control are proposed selectively through a closed feedback loop. By contacting with environment, the uncertainties higher. Self-tuning controller which adapts to variable dynamic response is applied to compliant motion control in order to satisfy the desired operation. The applicability of the suggested algorithm was confined by simulation of the contour tracking task of four joint manipulator.

There is an increasing interest in high-speed signal processing in modern telecommunication and consumer electronics applications. HDTV, ISDN. A limiting factor in Op-Amp based analog integrated circuits is the limited useful frequency range. This research program will develop a new CMOS Op-Amp architecture with improved gainband width product. The new design CMOS Op-Amp will achieve up to 100MHz unity gainband width with a 1.5-micron design rule.

In this paper, a CMOS Operational Transconductance Amplifier (OTA) which is used for high-frequency operation has been designed and simulated by SPICE 2G program. To increase input linear range, the input stage is designed by cross-coupled pair. And the output stage insert buffer stage for the buffing and gain. The band-width of designed OTA is $50{\sim}60$ [MHz].

Because Filters do both the sampling or Sampled-and-Hold(S/H) operations, it result in the replication of the spectrum of input signal and (unless Nyquist's criterion is satisfied) induce aliasing distortion. In this paper, therefore, we designed AAF(Analog Anti-aliasing Filers) so as to remove high-frequency components from input signal for aliasing distortion prevention before input signal had been processed frequency seletion by filters. and then, researched characteristics of it.

홉필드 신경회로망을 VLSI로 구현하는데 문제가 되는 area complexity를 개선한다. 적용한 응용은 layer assignment 문제이며, 기존의 layer assignment 신경회로망의 N ${\times}$ L개의 노드를 N ${\times}$ log L개로 감소시킴으로써 O( $(log\;L/L)^2$ )로 area complexity를 개선한다 (N : net의 수, L : layer의 수). 이를 위한 새로운 에너지 함수를 제안한다. 기존의 layer assignment 신경회로망과 동일하게 본래의 홉필드 신경회로망의 진동을 방지하기 위하여 self-feedback이 있는 수정된 홉필드 모델을 사용한다.

The main problem in speech recognition is the enormous variability in acoustic signals due to complex but predictable contextual effects. Especially in plosive consonants it is very difficult to find invariant cue due to various contextual effects, but humans use these contextual effects as helpful information in plosive consonant recognition. In this paper we experimented on three artificial neural net models for the recognition of plosive consonants. Neural Net Model I used "Multi-layer Perceptron ". Model II used a variation of the "Self-organizing Feature Map Model". And Model III used "Interactive and Competitive Model" to experiment contextual effects. The recognition experiment was performed on 9 Korean plosive consonants. We used VCV speech chains for the experiment on contextual effects. The speech chain consists of Korean plosive consonants /g, d, b, K, T, P, k, t, p/ (/ㄱ, ㄷ, ㅂ, ㄲ, ㄸ, ㅃ, ㅋ, ㅌ, ㅍ/) and eight Korean monothongs. The inputs to Neural Net Models were several temporal cues - duration of the silence, transition and vot -, and the extent of the VC formant transitions to the presence of voicing energy during closure, burst intensity, presence of asperation, amount of low frequency energy present at voicing onset, and CV formant transition extent from the acoustic signals. Model I showed about 55 - 67 %, Model II showed about 60%, and Model III showed about 67% recognition rate.

Hopfield crossbar assosiative network을 기초로한 개선된 Hopfield neural network을 제안하고, 이 network이 NP-complete 문제에 대한 효과적인 tool임을 보였다. 이 모델을 YLSI routing을 위한 layer assignment 문제에 응용하였고, 결과 이 개선된 Hopfield model이 stability와 accuracy를 향상시킴을 보여 주었다.

Texture provides an important source of information about the local orientation of visible surfaces. In this study the 3D shape of a textured surface is recovered from its perspective projection image on the assumption that the texture is homogeneously distributed. To recover 3D structure, the distorting effects of the perspective projection must be distinguished from properties of the texture. In this study, paraperspective projection, approximation of perspective projection, has employed.

This study introduces the three-dimensional information about moving objects. Relative motion between textured objects and observer generates a time varying optic array at the image, from which image motion of contours can be extracted. Closed-form solutions are proposed for the structure and motion of planar and curved surface patches. The analytic solution for curved surface patches combines the transformation of Longuet-Higgins with the planar surface solution of Subbarao and Waxman. Ovoid patches are shown to construct a unique transform angle. Thus, ovoid patches almost always yield a unique 3D interpretation.