1988.07

The UCERSS algorithm is an extended MUSIC which is used to estimate incident angles of multiple wide band signals. The purpose of this paper is to extend the UCERSS in order to estimate the direction of arrivals of multiple wide band signals in multipath environment. The modifications of the UCERSS result in the wide band spatial smoothing and the UNSS approaches. Computer simulation results indicate that the performances of the UNSS are superior to those of the UCERSS and the wide band spatial smoothing method.

Kmeans iteration method is generally used for creating the templates in speakerindependent isolatedword recognition system. In this paper the initialization method of initial centers is proposed. The concepts are sorting and trace segmentation. All the tokens are sorted and segmented by trace segmentation so that initial centers are decided. The performance of this method is evaluated by isolatedword recognition of Korean digits. The highest recognition rate is 97.6%.

An isolatedword recognition method using adaptively partitioned multisection codebooks is proposed. Each training utterance was divided into several sections according to its pattern extracted by labeling technique. For each pattern, reference codebooks were generated by clustering the training vectors of the same section. In recognition procedure, input speech was divided into the sections by the same method used in codebook generation procedure, and recognized to the reference word whose codebook represented the smallest average distortion. The proposed method was tested for 100 Korean words and attained recognition rate about 96 percent.

With the development of VLSI technology, research on special processors for highspeed processing is on the increase and studies are focused on designing VLSIoriented processors for signal processing. This paper processes a onedimensional systolic array for Discrete Hartley Transform implementation and also processes processing element which is well described for algorithm. The discrete Hartley Transform(DHT) is a realvalued transform closely related to the DFT of a realvalued sequence can be exploited to reduce both the storage and the computation requried to produce the transform of realvalued sequence to a realvalued spectrum while preserving some of the useful properties of the DFT is something preferred. Finally, the architecture of onedimensional 8point systolic array, the detailed diagram of PE, total time units concept on implementation this arrays, and modularity are described.

The musical tone of clarinet was synthesized by using nonlinear distortion method. In this method, the control of dynamic spectrum and harmonic contents can be easily attained. The method can be also applied to synthesis of musical tones of many other tonal instruments.

This paper proposed the phonemic segmentation method that employed two types of Likelihood Ratio that measures the change of spectral structure. By this method, isolated digits and words of VCV form are segmented into phonemeunit and especially, firstburst part in an aspirated bilabial plosive is divided.

A speech that play a part of important mediation in the man's conversation is the sound of representation to man's emotion and thought, then voice sound could be verified and identified a speaker's speech by individual property. This study indicates as distribution of pitch in searching for sample number of each pitch with eye in the sound waveform of speaker. We propose the algorithm that judge speaker's emotion state, personality, regional group, age, sex distinction, e.t.c., according to the deviation degree.

A new set of the units of demisyllables is presented for Korean speech synthesis. The performance of the set of demisyllable units is compared with that of the set of syllable units in the aspects of the quality of synthesized speech using each set of the units and the size of the computer memory which each set of units occupies. The set of demisyllable units achieves comparable speech quality and occupies smaller memory size than the set of syllable units.

In this paper, the starting point of affricate is detected from the first predictor coefficient of a 12pole linear predictive coding (LPC) analysis and phonemic segmentation is done through measuring short time energy and zero crossing rate. By this segmentation method, the duration of an aspirate can be mearsured in order to detect an aspirate or not.

In the continuous casting process, mold level sensor control system is very important for the quality of slab & bloom, in this paper, the principle of eddy current molr level sensor was studied and a method of mold oscillation signal detecting was purposed.

In speech signal processing, it is necessary to estimate exactly the pitch. We propose a new algorithm which uses the correlation coefficient between the original and the repeated segmental waveform in the frame as a parameter in the pitch extraction. The correlation coefficient in the frame reflects the periodic component and the transient ratio of the waveform.

In this paper, a methodology for determination of the thickness of layer on a substrate using reflection acoustic microscope has been presented. It has been shown that the amplitude and the phase of reflection coefficient of the layersubstrate composite has been used for measurement of layer thickness, acoustic velosity, mass density of the layer material. The reflection acoustic microscope operating at a frequency of 15 MHz has been used for the experiment and the measured acoustic impedance value for aluminum sample has agreed with the published data, and the measured layer thickness for silverglass composite has agreed with that measured using micrometer.

In this paper, Rayleigh wave velocity has been measured by detecting the backscattered signal generating near the Rayleigh critical angle in the elastic medium. The rotating system has been made for the measurment of Rayleigh angle. It has been shown that the measured results has been good agreement with the theoretical value. The method of nondestructive evaluation using backscattering signal has been presented and the internal of IC sample has been displayed acoustic image with good contrast.

In this paper, Digital signal processing method is deserved by the tissue characteristic quantization. Linear phase velocity is studied using linear signal, and the slope which attenuation characteristics are decreased is analysed according to frequency increase. A more efficient method determining the minimum phase function is developed from the spectral magnitud function.

The investigation of subsynchronous resonance by utilities has resulted in proposed and applied method to mitigate or eliminate SSRdamage to turbine generators. Many papers presented various methods and devices that can be used as countermeasures for the SSR. This paper introduced the availability of Superconducting magnet energy storage as an countermeasure for the SSR in the capacitor compensated transmission line.

This paper deals with some aspects for the design of induction servomotors with respect to permanent magnet servomotors and standard induction motors as control components. Induction servomotors have much lower weight and moment of inertia than the DC servomotor but cannot reach the excellent values of a brushless DC servomotor with rare earth magnets.

In this paper the application of variable structure control with sliding modes for improving the dynamic response and eliminating chattering of DC servo motor drive is presented. Sliding mode control is an effective means to keep a system insensitive to parameter variation and disturbances. In this method the control structure is changed discontinuously to farce the system dynamic to follow a predetermined trajectory. However the discontinuous change in control structure causes the controller input to chatter and gives nonzero steady state error. To overcome this problem, the discontinuous sign function is replaced by a proper continuous function, and presented through experiment.

The work is concerned with the analysis of the voltage escalation caused by the repeated restriking and extinguishing of the current when the vacuum circuit breaker interrupts the arc furnace load current. The paper particularly concentrates on the protective measures that may be adopted to overcome the restriking problem and guidelines are evaluated.

In this paper, the governing equation of the three phase linear induction motor is induced by the method that current in the primary winding is assumed as a current sheet. The steady state flux and the flux that is produced by end effect are mathematically analysed by electromagnetic equation. Each thrust of these flux is respectively induced. So in this paper, it is proved that linear motor is highly affected by end effect at high speed and speedthrust characteristics of three phase linear induction motor are shown by frequency variation in the actual model.

There are various causes for noise production in induction motors, and among them, the noise generated by magnetic vibrating distribution forces have an important meanings. In this paper, the magnetic vibrating distribution forces in squirrelcage induction motors were arranged systematically in the consideration of stator harmonics and it's forces were not considered the effection of airgap eccentricity and deformation. With the application of the finite element method in order to find magnetic flux density and with the calculation of the permeance of rotor node, the magnetic vibrating distribution forces were analyzed.

This paper proposed a new equivalent circuit and parameter estimation for I.M, which is different from T type and L type equivalent circuits. By using this circuit, we can analyze the torque of I.M, such as seperately exited D.C Motor, further more, we think that this equivalent circuit is effective to the vector control system for I.M.

A Study on the Improvement of Torque Characteristics of D.C Brushless Motor by Auxiliary Slot MethodIn all sorts of primary factor which produce torque ripple of d.c brushless motor cogging torque which is generated from relation between permanent magnet rotor and salient pole cannot controled by external control system. So in this paper cogging torque is analysed by simple method in which slot fuctuion is introduced. And for reducing counterplan of cogging torque d.c brushless motor which plot auxiliary slot by inequality interval is proposed. Thus the quantitative torque characteristics of new motor which designed by this method is analysed by F.E.M and validity of this new design is proved.

This paper reports a simple posteriori error estimate method for adaptive finite element mesh generation using quadratic shape function especially for the magnetic field problems. The elements of quadratic shape function have more precise solution than those of linear shape function. Therefore, the difference of two solutions gives error quantity. The method uses the magnetic flux density error as a basis for refinement. This estimator is tested on two dimensional problem which has singular points. The estimated error is always under estimated but in same order as exact error, and this method is much simpler and more convenient than other methods. The result shows that the adaptive mesh gives even better rate of convergence in global error than the uniform mesh.

The sliding mode control is an effective method to establish robustness against parameter variations and disturbance. But, in sliding mode strategy, the control function is discontinuous on the hyperplane. Consequently, the control input chatters at high frequency. When we apply such a control to the induction motor drive system, that causes a torque ripple and current harmonics, which are harmful to the system. In this paper, we introduce a low pass filter between sliding mode control output and driver input to overcome that problem. The bandwidth of this filter is varied according to the error funtion to improve transient responses.

The purpose of the paper is to operate a highdynamicperformance for ac servo system. It is describes a characteristics and stability regions by linearized control system of squarrielcage induction motor. Designed control system are compose to slipfrequency vector control for controlled a scandary interlinked flux by PWM inverter.

This paper is studied the new speed control system of the DC motor which can drive constant speed at variable load as well as reduce the homonics. System is used thyristorLeonard circuit with 2tap in order to reduce the homonics problems due to chopping dc input voltage in chopper system and semiconductor devices in thyristorLeonard system. There is added to the PLL system to drive the constant speed in variable Load, which has controlled of variable speed in DC motor. Thereby, we have proved its through experiment.

In this paper, a control system design method is proposed for DC motor drive. A state space model is used to control sysytem and for closed loop system the technique of pole assignment is applied. The control system is designed with state feedback theory and to improve the response further more feedforward theory is applied to control system. The microprocessor as a controller and the interfaces in the system are proposed. Digital simulation results for step changes in reference velocity and load torque are shown.

All parasitics such as switch conduction voltages, conduction resistances, switching times and ESR''s of capacitors are counted in the new statespace modeling based on nonideal switching functions. An equivalent simplified model is derived from the complex circuit with parasitics. Hence the results are very simple and exact, which are very important features of modelings. The pole frequency, dc voltage gain, and efficiency of the general converter, the buckboost converter are analyzed and verified by the experiments with good agreements with the theories. This may be a good summary for the previous works concerned with parasitics.

In this paper, We Proposed a new ThyristorFET hybrid type highFrequency inverter circuit topology. Concerning the proposed highfrequency inverter. We applied the normalized parameters to the characteristic analysis of steadystate at the resistive load.

A new current source inverter (CSl) with dcside commutation and loadside energy recovery circuit is proposed with analysis and explanation of the circuit operation. Proposed inverter overcomes the most drawbacks of the conventional CSI's  high device voltage stress, low operating frequency range, large commutation capacitance, etc.  by employing simultaneous recovery and commutation concept. The new CSI employs only one commutation capacitor and it can be built with considerably low cost. The commutation energies are temporarily stored into a large dc capacitor and recovered to the load side, thus the device voltage stress is low and the efficiency is high in the proposed inverter. Computer simulation results are given at the steady state, and a guideline determining the commutation circuit is given.

For the high speed drive of induction machine, Thyristor inverter utilizing line and device commutation is proposed. By investigating the load commutation conditions, the possible area of commutation is clearly shown. And from the above results, starting method under the constant slip frequency, is suggested.

The induction motor is constantly operated by general source, thus its speed control is employed an inverter system which can convert DC into AC. The CSI(Current Source Inverter) which have a commutation capacitor in its circuit is liable to cause a voltage spike that it is due to charge and discharge of commutation capacitor. And six phases inverter makes a number of harmonics. These have a effect upon the induction motor badly. This paper aims to suggest a way to reduce such adverse effects by maximally cutting the voltage spike as well as by eliminating a number of harmonics through the operation of MultiHFCSI.

Squarewave current source inverter drives suffer from torque pulsation and speed variation at low speed. Attempts to minimize these problems, CSI PWM strategies have been reported. It is shown that these PWM strategies are based on unnecessarily restrictive modulation laws. In this paper, trapezoidal PWM strategy for CSI is investigated theoratically by double fourier series and we proposed Harmonic Speed ripple Factor(HSF) that is independent of motor parameters and load conditions. Speed ripple are considered in TPWM and square wave inverter by HSF. We obtain modulation index(M) and carrier ratio (CR) for minimum speed ripple.

A usual linecommutated thyristor inverter with additional onebypass GTO is presented. With the added one bypass GTO, the extended commutating area is investigation and restrictions for the PWM to reduce the high frequency component of line current are shown.

In this paper, the characteristics of brushless DC Motor using voltage source inverter are analysized. The optimal commutation angle for Maximum Torque, the pulse width of supply voltage for available torque and the advanced commutation according to motor speed and electrical time constant effect power factor, efficiency, and no load speed limit.

A reactive power compensator whitch employs a three phase voltage source PWM converter is presented and analysed in this study. In this study, instantaneous reactive power compensator by the theory of instantaneous real and imaginary power are theoretically studied how to compensate reactive power occured by fundamental and harmonic reactive current. And we showed the compensation characteristics by comparing the experimental results with those of computer simulation.

In this study, so called high frequency cycloconverter which converts commercial so power source directly to high frequency power source is proposed. In this system, the circulating current play very important role. Its state and magnitude effects on system characteristics. This paper deals with the control method of the circulating current and side band component analysis of high frequency side current which is produced by it.

The currentfed DCDC converter, which is known as the most stable DCDC converter, has a twowinding reactor in series with the input. In this paper the new multioutput DCDC converter circuit, in which the 2nd winding of the reactor is creating the 2nd output, while the 2nd winding is feeding the energy to input in the currentfed converter, is proposed. The steady state characteristics of the new circuit are clarified. And it is found that the maximum value exists in the 2nd output.

The output characteristics of VSI by transistor depend upon bias signal with drive circuit. As the CTI is one of the VSI, takes case PWM output voltage. This Inverter operation is based upon position and negative of sinusoldal PWM signal by carrier triangular wave and compared sine one. Besides, the Inverter operation system required digital signal for microprocessor, sampling method of Inverter driver signals study for get rid of generated signal is error. In the paper, the method of 3 level PWM signal for CTI was studied. This method of signals depend upon caculated switching position of natural sampling with
$10^{6}$ modulation error for time flowchart programs. 
Quantum converters, a subset of resonant converters operating with optimal conditions are modeled. It is shown that series resonant converter(SRC) can be modeled as buck/boost converter with an equivalent inductor and parallel resonant converter(PRC) can be modeled as Cuk converter, with an equivalent capacitor. Also new resonant circuits with boost, buckboost and Cuk converter characteristics are proposed. From these models, the quantum converters can be designed to be controlled with closed loop feedback, having many advantages such as low device switching stress, reliable high frequency operation and low EMI.

This paper describes the development of the Interfacing unit between PSTN and interphone network. The designed and realised system detects the ring signal from PSTN and makes offhook condition when it is detected. Then the caller can hear the response tone from the unit and the tones of the interphone network condition. The developed system receives the interphone number to which the caller wants to connect by MFC tones and drives Interphone switch matrix. Thus, a call is connected to the interphone user directly. The caller may press the interphone number up to 10 digits and the last two of which are recognized as valid ones. The caller must press the first digit within 20 seconds after system response.

The purpose of this paper is to propose a simple firmware realization of PCM framing system, which exploits LSI memories for performance improvement and hardware simplification. The proposed system simply consists of a tapped delay line for simultaneous observation of in framingbits and ROM programmed sequence controller for framing process. Perpormance analyses are made in terms of misframe interval, synclossdetection time and reframe time. The proposed strategy proved to be significantly better in reframe time, stability and hardware implementation.

In this paper, the design of decoder for RS code using discrete finitefield Fourier transform is presented. An important ingredient of this design is a modified Euclid algorithm for computing the errorlocator polynomial. The computation of inverse elements is completely avoided in this modification of Euclid algorithm. This decoder is regular and simple, and naturally suitable for VLSI implementation.

This paper describes CODEC based on hybrid interframe and BTVQ including CRC(Conditional Replenishment Coding) system and RLC(Run Length Coding) method. Based upon the algorithm described in this paper, a practical CODEC has been designed for video conference and video phone use at 54Kbps using an ADSP2100 as a CPU.

In this paper, a codec for DECTED (31, 20)BCH code is realized. Moreover, using ROM and efficient elementary circuits in a decoder, we propose a method of making highspeed decoder.

In this paper, techniques of constructing parity check matrix of SBECDBED codes will be presented to improve reliability of mulitibitperchip type memory systems. And the high speed parallel CODEC of (88.64)SBECDBED code which is applicable to real system will be designed.

Since the bits in any encoded PCM word are of different importance to the bit positions, in order to improve the signal to noise ratio the technique that the encoded signal bits are weighted for the QPSK transmission system, is presented. Also the erasure zone is established at the detector, such that if the output falls into the erasure zone, the regenerated sample is replaced by interpolation. Two weighting methods are shown here. One is the method that the same weighting profile is used to Q and I dimension in QPSK signal constellations. The other is diferent weighting to Q and I dimension. The gains of this new technique in overall signal s/n compared to conventional QPSK transmission system were 5 db and 2db, respectively.

Some efficient methods for solving the equations over GF(
$2^m$ ) are proposed in this paper. Using these algorithms, parallel decoder for a tripleerrorcorrecting(31, 16) BCH code is implemented. By incorporating with ROM and PAL which are inserted in a decoder, the complex logic circuits can be substantially reduced and therefore a high speed decoder can be constructed. 
In transmission of DPCM signals, it is desired to protect the more significant digits from more errors than the less significant digits. The SNR of DPCM is examined in the case that bit error rates of individual digits consisting of the information word are different each other. The examination shows a better DPCM coding.

Lucifer  type algorithms with variable length of block size are presented. They are investigated and compared in the viewpoint of the intersymbol dependence. The results show that the Sbox and the number of rounds are very important factors in achieving the complete dependence.

The unequal biterrorprobability of rate r=b/n binary convolutional code is analyzed. The error protection affored each digit of the btuple information word can be different from that afforded other digit. The property of the unequal protection can be applied to transmitting sampled data in PCM system.

In digital mobile communications received speech data are affected by burst errors as well as random errors. To overcome these errors we propose a bitselective forward error correction scheme for the speech data which is subband coded at 13 kbps and transmitted over a 16 kbps channel. For a few error correcting codes the signaltonoise ratio of errorcorrected speech is obtained and compared through the simulation of mobile communication channels.

Direct Decoding Method for binary BCH codes which directly can find error location number from syndrome without calculating error locator polynomial is presented in this paper. The (255,239) BCH decoder is implemented using TTL logics. It is shown from our results that this decoder can be implemented with relatively simple hardware.

This paper presents a new uniform sampling method for twodimensional contour. The proposed method improves the uniformity of contour sampling in terms of the interpolation of the corner points extracted from chain codes. The performance of the proposed method is analyzed through the length estimation and shape recognition of twodimensional contour. The experimental results show the significant improvement over the conventional method.

This paper describes design & fabrication of RF modulator using 2um Bipolar process which convert video & audio signal into high frequency VHF/UHF signals for all TV standards. This VHF/UHF RF modulator fabricated using 2um bipolar process( T max = 5GHz) shows satisfying electrical characteristics and meets all the design targets.

Preprocessing is indispensable to eliminate local granularities prior to region growing. In this paper, we examined the effects of preprocessing in S&M region growing technique. Experimental results show that a modified Nagao filter removes the local granularities well and compensates for the defects of Nagao filter.

In this paper, the transform coding method was selected out of image compression techniques. Each characteristic of transform coefficients for five transform(DCT, DST, WHT, HAT, SLT) was observed, and their performances were compared and reviewed, based on the results from the experiment where image samples were applied to five transform. As the result of image processing experiment, DCT was found to have the best performance.

A new transmission scheme is presented for transmitting motion video at very low data rates. In this scheme, twolevel cartoons are extracted from graylevel images, and their interframe differences are coded using 2D runlength coding. Experimental results show that the proposed scheme yields compression ratio as high as 216:1.

In this paper, we design adaptive DPCM by the two dimensional linear prediction using the covariance method and propose adaptive algorithm which determines analysis frame size by the compassion of the local standard deviation and the global standard deviation of images.

Recently, an iterative reconstructionreprojection (IRR) algorithm has been suggested for application to incomplete data computed tomography (CT). In the IRR, the interpolation operation is performed in the image space during reconstructionreprojection. The errors associated with the interpolation degrade the reconstructed image and may cause divergence unless a large number of rays is used. In this paper, we propose an improved IRR algorithm which eliminates the need for interpolation. The proposed algorithm adopts a new sampling scheme in which samples (projection data) is taken in phase with the samples of the Cartesian grid.

Prediction errors generated by motion compensated coding are coded with transform coding techniques as DCT. The performance of transform coding techniques are lowered mainly due to the source characteristics with a great deal of zero populations and plusminus sign changes, i.e., low correlation. In this paper a transform coding scheme which adopts a decomposition of prediciton errors into two sources is proposed and compared its performance with conventional scheme.

This paper proposes an improved ART (Algebraic Reconstruction technique) algorithm. This algorithm is an iterative one with iteration sequence criteria based on the discrepancy between measurement and pseudoprojection data. The simulation result using the proposed algorithm shows a significant improvement in convergency rate over the conventional ART algorithm.

A lot of time is required in order to the process of loop repeating the preprocessing operated in the software. Specially in the preprocessing, most of the time is used for the noise elimination such a software algorithm component of a noise elimination hardware, this can operate quickly the process.

The most spaceconsuming element of digital image processing system is the video memory. Though this problem is solved by DRAMs, timing constraints posed by video data rates. The cycle time of DRAMs can be diminished by serial transferring and reading or writing pixel datas at the same time. This paper resents the design of 1024
${\times}$ 512 video memory using this technique. 
원자력발전소의 운전현황감시 및 안전성확보를 위해 원전에 비상대응설비등을 설치 할려고 하는 상황에서 본 연구는 이런 원전설비 시스템의 입력단이 되는 DAS(Data acquisition System)를 설계, 제작하므로 원전설비에 대한 기술자립을 향상시켰다. 여기서 연구된 DAS는 현장신호를 격리 및 분리하고 analog와 digital 신호를 취득하며 Non Class1E analog 80개, digital 신호 16개와 Class1E analog 신호 2개, digital 신호 4개를 취급할수있도록 설계하였다. 여기서 설계된 DAS는 원전의 안전정보 system이나 기타 process의 monitoring system의 입력단으로 적용 가능한 시스템이다.

A protocol on transferring informations between IBMPC/AT and STC(self tuning controller) by using the RS485 bus is suggested. The operating algorithm is designed so that the host computer monitors the status of STC, and can be updating parameters depending on the environmental conditions for the various oprating mode.

In this paper, a transmission protocol including error recovery strategy on the data link layer for Ethernet using CSMA/CD media accessing method was proposed. So when considering the actual transmission error probability on the channel, it's performance was analyzed through a simulation. Performming the simmulation, the required parameters was taken as those given by Ethernet controller interface board.

Several routing algorithms use the link lengths that reflect the traffic conditions on the links to route messages. The idea is to assign a large length to a congested link so that they will tend not to use it as part of a routing path. When these lengths depend on current traffic conditions, dynamic behavior such as stability, convergence and the speed of convergence are of interest. But it sometimes induces the possibility of violent oscillation. This paper is the one that analyze these issues systematically.

This paper deals with the test strategy on the shortopen for the printed circuit board. A group testing algorithm, which is the several testing point to be measured redefined as one of the testing points, was suggested. As a result, the total testing time was reduced to 30
${\sim}$ 50 percent. 
This paper is to show design of robot manipulator which has 3link using DC Motor and realization of control algorithm with IBM  XT Microcomputer connected. Gentral algorithm is applicated by position and pass control using pointtopoint method. At first, this paper computes required angles on each joint in order to search desired position or path, and uses a voltage control with feedback from output of encoder and tachometer in real time. The application of control algorithm on position, velocity and force for each joint of manipulator by using selftuning control is left for next study.

This paper realizes the multioutput truncated difference circuits using current mode CMOS, and presents the algorithm designing multi  valued logic functions of a given multivalued truth tables. This algorithm divides the discrete valued functions and the interval functions, and transforms them into the truncated difference functions. The transformed functions are realized by current mode CMOS. The technique presented here is applied to MOD4 addition circuit and GF(4) multiplication circuit.

In this paper, a method for constructing parallelin, parallelout multipliers in GF(
$2^{m}$ ) is presented. The proposed system is composed of two operational parts by using shift register. One is a multiplicative arithmetical operation part capable of the multiplicative arithmetic and modulo 2 operation to all product terms with the same degree. And the other is an irreducible polynomial operation part to outputs from the multiplicative arithmetical operation part. Since the total hardware is linearly m dependant to an GF($2^{m}$ ), this system has a reasonable merit when m increases. And also this system is suited for VLSI implementation due to simple, regular, and concurrent properties. 
This paper suggests the way to improve the impulse response characteristics in 1D and 2D FIR digital filters design. The proposed a Least Square(
$L_2$ ) error algorithm of kaiser window is the better than$L_2$ error of the Remez Exchange(R.E.)algorithm. 
This paper shows the development of robot sensor system using optical fiber. When the object recognition is implemented in the existing robot system, the image processing method using CCTVCamera has been most widely used. But when it is necessary to recognize and classify only simple elements in industrial field, the real time processing using this method requires relatively high cost. The purpose of this paper is to represent the fundamental study on the development of opticalfiber sensor system, which may be used to recognize and classify elements with low cost and realtime.

The tomographic imaging that employs ultrasonic echos has achieved outstanding advances in recent years, and today, ultrasonic diagnostic equipment has become the tool that is absolutely indispensible for clinical operations. Meanwhile, the feasility of measuring blood flow in the heart and vessels by the use of Doppler effect in ultrasonic waves is a well known fact. With respect to the method of blood flow measurment, there are two kinds which employ continous wave and pulse wave doppler system. In this paper, we describe the measurment of Blood flowturbulence using general purpose Digital Signal Processing Board which had been implemented for the purpose of realtime spectrum analyser. Blood flowturbulence means the bloodflow behavior. And it's value proportional to the spectrum variance. Therefore mean frequency of blood signal and variance provide useful diagnostic information. We have applied to the major arteries and vein, obtained the information about the time dependent bloodflow behavior.

The ultrasound Doppler effect is used for measuring the velocity of the bloodflow in an artry. Because of the range information, the pulsed doppler system is most commonly used. In this paper, we propose a new pulsed doppler system which uses a quadrature sampling method in R.F. range in order to detect the bloodflow direction and to simplify the compexity of hardware. The pulsed doppler system using quadrature sampling method in R.F. range eliminates Inphase, Quadrature phase channel balancing problem at demodulator. In addition, the improved pulsed Doppler system shows the possibility of serial processing.

A pHmeter has been newly developed by using on ISFET as the ion sensing element. The performance characteristics of the developed pHmeter have been investigated and confirmed to be used practically. The stable operation and the temperature compensation were stressed in this study.

Pole assignment controller with variable forgetting factor is generalizaed to allow the output and/or input variance to be reduced. The algorithm can give significant reductions in variance for little extra computational effort and is presented for servotracking using leatsquares estimation. Moreover, the use of a variable forgetting factor with correct choice of information bound can avoid 'blowingup' of the covariance matrix of the estimates and subsequent unstable control.

This paper describes 256k (256K
${\times}$ 1) CMOS SRAM utilizing 1.2um doublepolysilicon and doublemetal CMOS process. A typical access time of 16ns with a 30pF load has been achieved through the use of a block architecture, a new decoder, an unique bitline scheme and an optimized process. Operating current is 55mA at 40MHz and 15mA at 10MHz. A highresistive polysilicon load has been used to achieve a standby current of 3uA. 
PHOTORESIST 자체문제로인해 감소되는 최소해상력, 촛집심도여유 및 CRITICAL DIMENSION (C.D.) 조정여유도등을 연구하였다. 기존에 사용중인 PHOTORESIST(큰 분자량)와 PHOTORESIST자체내에 CONTRAST 촉진 물질(CEM)이 첨가된것(INNER CEM TYPE) 및 PHOTORESIST구성성분중 작은 분자량/좁은 분자량 산포가 형성된 RESIN 의 PHOTORESIST(LOW MOLECULAR WEIGHT CONTROL TYPE)등 세가지 PHOTORESIST를 사용 하여 상기의 항목을 분석하였다. INNER CEM TYPE 및 LOW MOLECULAR WEIGHT CONTROL TYPE의 PHOTORESIST는 기존에 사용중인 RESIST보다, 최소 RESOLUTION은 약 0.2  0.3 um, DEPOCUS MARGIN은 약 0.8  1.2 um 및 C.D. CONTROL LATITUDE 향상된 것 등이 우수하였다.

For finding the effects of As on
$TiSi_2$ formation, sputter deposited Ti film on Si substrates implanted with various doses of As have been rapid thermal annealed in Ar atmosphere at temperatures of 600900$^{\circ}C$ for 20 sec. The sheet resistance of TiSilicides was examined with 4point probe, the thickness with${\alpha}$ step, and the As dopant behavior in Si substrates with ASR. The thickness of TiSilicides decreased with increasing As doping, but TiSilicides sheet resistance increased with increasing it. However, the critical concentration effect reported by Park et al. was not observed. We observed that the thickness of native oxide increase with increasing As doping. Thus, we concluded that native oxide act as a "barrier" for the Si diffusion. 
The photovoltaic performances of asi : H
$n^{+}pp^{+}$ solar cells have been investigated. The optimum substrate temperature for the deposition of aSi : H$n^{+}pp^{+}$ cell decreases with increasing doping concentration of the player, and is less than 200$^{\circ}C$ when the gas phase doping concentration is higher than 10 ppm. The results can be explained as the dependences of substrate temperature for the relaxation of silicon atoms and for the bonded hydrogen concentration in the player. 
Sealed Interface Local Oxidation (SILO) technology has been investigated using a nitride/oxide/nitride threelayered sandwich structure. Ptype silicon substrate was either nitrided by rapid thermal processing, or silicon nitride was deposited by LPCVD method. A threelayered sandwich structure was patterned either by reactive ion etch (RIE) mode or by plasma mode. Sacrificial oxidation conditions were also varied. Physical characterization such as crosssection analysis of field oxide, and electrical characterization such as gate oxide integrity, junction leakage and transistor behavior were carried out. It was found that bird's beak was nearly zero or below 0.1um, and the junction leakages in plasma mode were low compared to devices of the same geometry patterned in RIE mode, and gate oxide integrity and transistor behavior were comparable. Conclusively, SILO process is compatible with conventional local oxidation process.

Influence of incoherent lamp heating sources on the absorption effect and variation of OSF' size were investigated. The absorption effect on I.R lamp caused by free carrier excitation is greater than that of TungstenHalogen lamp. The variation of DSF' size weakly affected by oxidation time.

A numerical method is presented for the scattering by the perfectly conducting cylinder with arbitrary cross sections. The relevant integral equation considered by the Efield formulation is solved by method of moments, and thereby the surface current induced as well as the radar cross section of the scatterer are numerically computed to specify the scattering nature of the scatterer. Two separate methods, one with point matching and the other Galerkin's method, are considered to make cross checks to the results obtained. Taking two half pulses suggested to expand the surface current shows savings in computation time and accurate solutions for the corners on the scatterer.

Typical magnetic bubble propagation failure modes of ion implanted magnetic bubble computer memory devices were observed and their failure mechanisms were analize. The skidding failure mode is due to the pushing of a strong repulsive charged wall. If this pushing is stronger than the edge affinity of the bubble in the cusp, the bubble moves out of the cusp when it is supposed to stay there. The stripeout failure modes across the adjacent track or along the track can be explained by considering the relative strength of the charged wall and the edge affinity encountered by both ends of the stripe. The skipping of the first cusp of a track is believed to be due to the whipping motion of the charged wall. The bubble moves directly to the second cusp via the long charged wall pointing to the second cusp skipping the first cusp.

In this paper the fabrication and application of CompositeThermistor were studied. CompositeThermistor specimens were made by conventional process. Some compounds of vanadium oxides and semiconductor oxides as a fillers were used to fabricate CompositeThermistor, and Epoxy Resins are used as a polymer matrix. The results of resistivitytemperature characteristics were measured in the range from 100 (
$^{\circ}C$ ) to 200($^{\circ}C$ ). The harder polymer, Eccogel series #136580 and Spurr Epoxy are more preferable compared to the Eccogel (136545) with some fillers. 
Conductance technique is the moat accurate method and gives more detailed information about interface of the MIS structure than other methods. With the measurement of the equivalent parallel conductance and capacitance, the characterization of SiSiN interface is developed. The interface state density of SiSiN is obtained by
$8{\times}10^{11}$ $6{\times}10^{12}(eV^{1}cm^{2}$ ). After the positive BT stress is performed on the sample, the interface state density gets increased. The interface state density is not effected by the D.C. stress. 
The properites of Zndiffusion in IIIV ternary alloy semiconductor
$In_{1x}Ga_{x}p$ , which was grown by the temperature gradient solution (TGS) method, have been investigated. The composition, x, dependence of the Zndiffusion coefficient at$850^{\circ}C$ and the activation energy for Zndiffusion into$In_{1x}Ga_{x}p$ were found to be$D850^{\circ}C$ (x)=$3.935{\times}10^{8}exp(6.84{\cdot}x)$ , and$E_{A}(x)=1,28+2,38{\cdot}x$ , respectively. From this study, we confirm that the Zndiffusion in$In_{1x}Ga_{x}p$ was explainable with the diffusion mechanisms of the interstitialsubstitutional, which was widely accepted mechanisms in the IIIV binary semiconductors. 
An FET type urea sensing device has been developed using the ISFET and platinum reference electrode and by employing the differential amplification. The performance characteristics of the fabricated sensor show a good linearity in the urea concentration range of
$5{\times}10^{5}$ to$10^{3}$ g/ml and a stable response. 
The new oxidation method was presented to grow the oxide layer by thermal reaction of
$NH_{3}$ and$O_{2}$ . The growth rate increased according as increase of partial pressure of$NH_{3}$ . Optical transparent of the grown film was 12% compared with 17% of thermal oxidation when the wave number was$1,100cm^{1}$ . The oxide layer with good quaility was obtained. 
A new
$NH_{3}O_{2}$ oxidation method was estimated by the electrical properties of the fabricated nMOS transistor. For the CV characteristic curves the Qox are almost equal to Qss and no hysteresis phenomena are observed. The IdVds characteristics show that$NH_{3}O_{2}$ oxidation method is superior to Dry oxidation. 
We suggest a new annealing method for stabilization of
$SiSiO_2$ interface state in MOS device using$NH_3$ (10%) +$N_2$ (90%) ambient gases. The annealing effect was examined through CV characteristics, threshold voltage, effective mobility on channel, respectively. The experimental result show that the new method is available to improvement of MOS device characteristics. 
This paper is concerned with the investigation of the impurity redistribution process in a two step diffusion. In integrated circuit technology, two step boron diffusion involving a deposition step followed by a drivein step in commonly encounted. The drivein process is usually performed in oxidizing atmosphere resulting in redistribution of impurity (boron) within the semiconductor. This paper proposes a new numerical analysis method; Bounary Element Method to determine impurity profile at the arbitrary point in domain by its coordinate and boundary value.

In this paper a newly designed rapid thermal process (RTP) structure is proposed to the slip induced in silicon wafers considerably. The reflectors and a graphite radiation were used to compensate the temperature difference causing slip in silicon wafers. From our experiments it is known that slip can be removed during a rapid thermal annealing at high temperature.

In this paper optimal light source arraies are calculated to reduce slips in RTA process. A twochannel temperature controller is constructed on a board using IBM  XT to improve the temperature uniformity. The proposed RTA structure has also advantage of power dissipation.

In MOS integrated circuits, annealing after oxidation process is necessary to improve physical properties of silicon dioxide. With subsequent annealing in inert gases such as nitrogen or argon, and excess silicon bond is allowed time to complete the oxidation and surface charge density is reduced. In this paper, we will present effects of the rapid thermal annealing on silicon dioxide. In order to evaluate characteristics of silicon dioxide, we analyzed CV curve dependent on annealing time and temperature, and presented variation of fixed oxide charge.

The increases in the packing density and the resulting shrinkage of silicon integrated circuit dimensions led to the investigation and successful of the deposited silicide layers as the gate and interconnection and contact metallization. In this paper evaporated Co films on nSi have been rapid thermal annealed in
$N_2$ ambient at temperature of$400^{\circ}C1000^{\circ}C$ . The Co silicide formation is characterized by sheet resistance (4PP). Also, silicide growth rate and its reproductivity has been examined by SEM. 
In discharge excited excimer laser, it is hard to say that the analysis of laser operation was well explained until now. But this can be improved by analysis the nonlinear discharge characteristics in the cavity. The nonlinear characteristics can be analysed by solving the nonlinear resistance which depends on electron mobility and number density. We can calculate the electron mobility and number density each other using Boltzmann equation and Kinetics equation. So we calculated the nonlinear resistance and analysed nonlinear discharge characteristics.

We report on the simulation results of 1)the second harmonic conversion with
$1.053{\mu}m$ Nd:glass laser using monolithic KDP crystal and 2)the efficient conversion from$1.053{\mu}m$ to$0.35{\mu}m$ by the polarizationmismatch 3rd harmonic conversion in two TypeII crystals. 
This paper describes the experiment and design of Rogowski  wound coil used to measure a pulsed current with pulsewidth on the order of 100nsec in the kiloampere range. Radius of coil is 2.75cm, total number of turns are 26 and cross sectional area is 1 cm. The coil have 3ns rise time, 0.2nsec transit time and sensitivity is 4v/kA.

With dichromated bleacher and alcohol drying, over effi. of 81% has been achieved for holographic gratings. It is recorded by two plane waves at 632.8 nm in Agfa 8E75 film. Swollen emulsion and rapid dehydration make a strong modulation. The model works well close to the Bragg condition. It is suggested modified parameters of coupledwave theory which have good agreement with experimental results.

For the efficient recognition of Korean alphabets, modified ternaryphase synthetic discriminant function filter with reduced resolution requirements, quantized phase levels and capable of implementing distortion tolerant pattern recognition is proposed.

In this paper, the realtime optical associative memory using multiple hologram which is generated with two angular multiplexed reference beams and Fourier transformed object beam in the
$BaTiO_{3}$ crystal based on DFWM mechanism. When one image is recorded in the$BaTiO_{3}$ crystal, complete image can be recalled by 9 % partial input of the stored original image without any additional thresholding and optical feedback process. As an experimental result of multiple Fourier hologram which is recorded with two binary images, OHCHAS and PARKHK, we can obtain complete image recalled by 1/6 partial input of the stored image. 
신경회로를 모방한 병렬 아날로그디지탈 변환기를 설계하고 광학적으로 구현하였다. 이 회로의 동작 원리는 주어진 입력 아날로그 값에 대해 출력 bit 들의 디지털 값을 동시에 추정하는 것으로
$2^{N}$ 단계의 구분을 위해서 N개의 단위 소자가 필요하다. 에너지 최소화 방법에 의해 설계된 신경망 아날로그디지털 변환기와 비교해 볼 때 회로의 구조가 단순하고 출력이 회로의 초기 상태에 관계없이 주어진 입력에 의해 결정된다. 
Holographic interferometry is applied in dental research to study the motion of teeth and the deformation of the associated facial structure. We consider double exposure holographic interferometry which has limiting factors such as the stability of fringes, the recording time, the amount of exposure, the localization of fringes and sensitivity. These factors are taken into account to make interferogram. The resulting interferogram gives fringe pattern which is analyzed qualitatively to determine the relative rotation and the deformation of the facial structure.

The structure consisting of an Eplane dielectric slab partially filling a rectangular waveguide is examined with attention on those higher order mode propagation characteristics that are relevant to the design of nonreciprocal remanence ferrite phase shifters. The mechanisms for the elimination of
$LSE_{11}$ ,$LSM_{11}$ ,$LSE_{12}$ modes are introduced. Experimental verification of elimination of higher order mode is shown in the form of insertion loss measurements. 
A simple design method of a single balanced MMIC mixer is described. It uses small signal S11 and capacitive load for the input matching circuit and the output loading circuit, respectively. It is found that the conversion gain of the FET mixer is independent of FET gate width. The fabricated mixer has 2.5 dB conversion gain at 9 GHz with 50 ohm IF load and 2 dBm local oscillator power.

Radiation admittance of the rectangular microstrip antenna is calculated by using equivalent
${\pi}$ network parameters of the slit in the wall of the parallel plate waveguide filled with homogeneous dielectric. Using equivalent transmission line model, return loss is calculated and compared with experimental result. Calculated values for the return loss show good agreement with experimental values. 
Rectangular microstrip antenna with parasitic element is analyzed. Radiation admittance and equivalent circuit parameters between rectangular microstrip antenna and parasitic element are obtained by using equivalent
${\pi}$ network parameters of the slit in the wall of the parallel plate waveguide filled with homogeneous dielectric. The return loss is calculated and measured as a function of frequency with gap width 0.5mm between the patch and parasitic element. The experimental results are in fairly agreement with calculated values. 
The purpose of this study is to evaluate the discontinuity effects of two kinds of transmission lines. i.e. the strip line and the microstrip line which are popular transmission structures.

Computation of induced currents on a resistive strip by the iteration method in the spectral domain.The induced currents on a uniform resistive strip by a Hpolarized plane wave are obtained by the iteration method in the spectral domain and compared with the results by the moment method.

In this paper, the reconstruction of complex permittivity distribution on a rectangular cross section of inhomogeneous dielectric cylinders is performed by employing the spectral inverse scattering scheme. Numerical simulations provide the superresolution to the permittivity profiles nearly regardless of the measurement locations of the scattered field and the permittivity distributions on the cross section.

A spectral inverse technique, which was developed by applying the pulse basis moment method procedure on the direct scattering problem in the reverse sequence for the reconstruction of complex permittivity profiles inside inhomogeneous dielectric objects, is modified to be applicable to the moment method with seriesexpanded basis. By performing numerical simulations for various type of dielectric objects, it is demonstrated that this inverse technique provides close reconstruction of permittivity profiles. Futhermore, compared to the previous scheme of the pulse basis, the presented method is shown to reduce the computation cost, relative error of reconstructed permittivity profiles by averaging in each cell, and the illposedness inherent to this inverse scattering problem.

A numerical scheme is considered for computing charge distributions and drawing equipotential lines for multiconductor transmission lines with arbitrary cross sections. Twodimensional integral equation superposed by the charges is numerically calculated by the method of moments. Here in Galerkin's method special pulses cut into halves chosen for expansion and testing functions facilitate the numerical computations in solving the integral equation. Numerical examples are given for some system of which the shell conductor is grounded as the reference zero of potential.

We present here a numerical solution for obtaining propagation characteristics including lossess for various modes of an multilayered optical waveguide structure. The method is based on transmission line analogy. A comparison with other method shows the our results are accurate and simple.

In order to use mBIZ code as the practical transmission code, the information block length have to be determined. If the information block length is short, timing information disapperance is reduced, but transmission speed is increased as the result of inserted redundancy bit. While, if the information block length is long, timing information disapperance and jitter are caused. So it is very important to determine the optimal timing information block length. Therefore, in this paper, we determined the optimal information block length using S/X endurance method.

The output power is dependent of the vibrational level temperatures and wall temperature of the discharge tube in cw
$CO_{2}$ lasers. The method postulates the introduction of a vibrational temperatures Ti for each vibrational mode. The vibrational and wall temperature are dertermined by the equations of the vibrational energy balance and thermal conductivity. 
The basic modeling is analyzed on the optoelectronic properties of CSPDH laser structure using selfconsistent calculation of optical field and the electronhole distribution in the active region. Laser properties is modelled include gain profile, threshold, near field and far field pattern. This new characterization is allowed for consideration such as carrier spatial hole burning due to strong optical fields which stimulate recombination.

As to the development of the information society, the common communication network which processes the service for data, image and voice etc., is required. So the higher degree of reliability becomes more important. Therefore, this paper describes the improvement of reliability using the format transformation of the mBIZ code and the reframe usually practiced at the terminal repeater.

A new method of optical short pulse generation using stable diode lasers with negative optoelectronic feedback is demonstrated experimentally. About 1 nanosecond pulses with 300 MHz repetition rate are generated.

Linkbylink and endtoend signaling affect the probability of inefficient trunk seizure, the usage of backward signal features, post dialing delay, total signal information processing load, each differently. Considering the circumtances of the domestic public switched telephone network, we conclude that endtoend signaling is better suited to the domestic network.

This paper describes the universaltype signaling service equipment which provides a signaling service function (R2 MFC, DTMF, Continuity check tone, and Audible tone sending and/or receiving) for call processing control in TDX10 switching system. Based upon the structure of the TDX10 switching system, the philosophy of design and technology on the universaltype signaling service equipment has been investigated in order to realize a reduction in size and cost, as well as assuring high reliability and flexible operation. Finally, the control method, fuctional operation and principal feature of the designed universaltype signaling service equipment have been described.

As computer systems get complex more and more, the powerful but easytouse man machine interfaces are required. This paper describes the man machine interface in the TDX10 switching system currently being developed at ETRI. The man machine interface in TDX10 provides a reliable and userfriendly system interface through which system operators can manage the system in a convenient way. It incorporates the window, menu, form, online help, history recording, command file batching and color graphics capabilities.

In this study we consider three types of accomodation methods of narrowband/wideband services. The fully integrated accomodation method is sufficient to accomodate the relatively low wideband traffic and the partially integrated method is required to meet the specification of blocking probability as the wideband traffic increases. The segregated accomodation method is required expecilly in case of the high wideband traffic demand.

This paper is to demonstrate the effectiveness of one point grounding in support of ESS protection. One point grounding provides for the dispatching of excess energy for protection of switching equipment and personnel from lightning discharge hazards and for a natural sink for noise from atmospheric lightning and power transients. In most ESS installations there are a number of different items that must be maintained at a common ground potential for safety, fault protection or noise reduction. The items typically consist of power systems, heating and ventilating systems, distributing frames, repeating equipmets, switching equipments, etc. Grounding system of an ESS Office is most effective when all the ground points are connected to a single, common earth. The one point grounding prvides a common reference potential, keeping all the items of telecommunications facility free from the earth current and voltage hazards.

This paper describes an effective knowledge acquisition method in the fault diagnosis expert system for the electronic switching system TDX1A. The knowledge acquisition procedure consists of the knowledge collection, the fault diagnosis modeling and the knowledge representation. Furthermore, to improve the performance of the knowledge bases with rule and frame representation, we showed the knowledge base checking methodlogy by using redundancy and inconsistency check algorithm.

This paper describes an effective trunk loop signal processing method and analyzes execution time of program in the DTCP(Digital Trunk Circuit Processor) in the TDX1A digital switching system. To predict a maximum trunk capacity, also analyzes to Z80A system clock(4Mbit/s, 2.5Mbit/s) and scanning period(8mS,5mS) respectively.

The switching system and transport will be developed and serve as hybrid switching system and link respectively according to the needs of mixed voice and data service in ISDN era. This paper describes a theory of optimal band width allocation in multiplexing voice and nonvoice traffic, and analyzes traffic performances on a model network.

This paper describes BSDL(Behavioral/Structural Description Language), CDTF(Control Data Text File) and state synthesizer built for use in ICSC(InCheon Silicon Compiler). BSDL describes structral and behaviral specifications of an ASIC(Application Specific IC) for digital system design. ICSC's paser generates CDTF consists of ifthenelse, arithmetic and data transfer statement according to each BSDL statement. State synthesizer generates CCG(Control Constraint Graph) in consideration of execution of statement and generates VCG (Variable Constraint Graph) in consideration use of variable generation and use of variable. Also, it involves allocating algorithm operation nodes in the data path and the control path to machine states with minimum state number and as small area as possible.

In this paper, a new multioutput logic minimization algorithm is presented. A base minterm is selected in the given function and the prime implicant is obtained by expanding it in the order of the expansion set that is decided by heuristic method. Inputoriented expansion procedure is used to reduce fanin and fanout number. To show the effectiveness of this algorithm, comparative run time with other minimization algorithm is given.

This paper proposes a graph matching algorithm based on simulated annealing, which assures the globally optimal solution for circuit partitioning for the placement in the rectilinear region occurring as a result of the preplacement of some macro cells, or onto the nonplanar surface in some military or space applications. The circuit graph (
$G_{C}$ ) denoting the circuit topology is formed by a hierarchical bottomup clustering of cells, while another graph called region graph ($G_{R}$ ) represents the geometry of a planar rectilinear region or a nonplanar surface for circuit placement. Finding the optimal manytoone vertex mapping function from$G_{C}$ to$G_{R}$ , such that the total mismatch cost between two graphs is minimal, is a combinatorial optimization problem which was solved in this work for various examples using simulated annealing. 
For CMOS, this paper propose a new global routing algorithm in which macrocells and I/O padcells can be treated. Not only predefined feedthrough in base array, but also some polysilicon line which are not assigned as inputs are used to prevent the overflow of nets passing through the row. The signal nets are assigned on their feedthrough by the maze router. By treating macrocells and I/O padcell, the routing from internal to I/O cell can be done automatically and a kind of is constraints in design process can be reduced.

This paper presents a new shadow testing acceleration scheme for ray tracing called hybrid shadow testing(HST), based on conditional switching between the conventional shadow testing method and Crow's shadow volume method, with the object polygons as well as the shadow polygons registered onto the corresponding cells under the 3D space subdivision environment. Despite the preprocessing time for the generation and registration of the shadow polygons, the total shadow testing time of the new algorithm was approximately 50 % of that of the conventional shadow testing method for several examples while the total ray tracing time was typically reduced by 30% from the conventional approach.

EDAS_P system for electronic circuits design, which is developed at ETRI. There are 3 kinds of graphics tools. In this paper, we presents the improved graphics functions of EDAS_P system using the GSS
${\times}$ CGI. 
Lee, Won;Chung, GabJung;Kang, JongHoon;Kweon, GyuWan;Kim, DaeHwan;Chung, HoSun;Lee, WuIl 545
This paper describes the advanced KUIC_LED which is a 2dimensional graphics editor for IC mask layout. This system runs on IBM PC/AT with the${\Omega}$ /PC graphics board. Though this system runs on PC, it offers about fifty commands to do most kinds of layout and 16 mask layer. KUIC_LED(${\Omega}$ /PC_VERSION) is written in C language. 
Jeong, GabJung;Lee, Won;Kweon, GyuWan;JongHoon, Kang;DaeHwan, Kim;HoSun, Chung;WuIl, Lee 549
KUIC_LED II is a two dimensional graphics editor for IC mask layout which breaks through the memory limit, maximum number of box is 2333, with overlay. It runs on IBM PC/AT with the${\Omega}$ /PC color graphics board. I/O data format is CIF(Caltech Intermediate Form). It is written, In 'C' language on MS_DOS. 
Recognition of IC chip pattern requires extraction of features, which have the information of vertex position and orientation. Edges are extracted and straightening algorithm is applied to the edges, so that lines are obtained. With these extracted data, the coordinate and orientation of all vertices are extracted and vertexform matching is applied to the locally overlapped area of neighborhood frames to have global coordinate of IC chip.

In this paper, we present a new kway partitioning algorithm for a graph of an electrical circuit wherein nodes and edges are regarded as cells (modules) and nets, respectively. In contrast to the previous work, our method is based upon a linearly ordered partition paradigm. We also claim that the maximum number of netcuts mostly governs the performance of kway partitioning, thus having influence on the construction of a new cost function. In addition, our approach elaborates upon balancing the partition size. Our experiments show excellent results in comparison with previous kway partitioning algorithms.

A currentcontrolled CMOS operational transconductance amplifier(OTA), whose transconductance is directly proportional to the DC bias current, has been developed for many electronic circuit applications. It features that its transconductance is insensitive to temperature unlike that of the bipolar OTA. This property makes it possible to use the proposed OTA as a basic buliding block in electrically variable circuit design. The SPICE simulation shows that the conversion sensitivity of the circuit is 44.62 mv /
${\mu}A$ and the linearity error less than 0.54 % over a bias current range from 2${\mu}A$ to 120${\mu}A$ when the output is loaded with a 1${\Omega}$ resistor. 
This paper describes the design of FIR filtering DSPchip that can be operated without programming. The proposed DSPchip has not only the improvement of execution time but also selectivity of filter length from N=1 to N=128. Hilbert Transformer can be designed from this chip. FIR filter system is composed of Data memory/Control Unit, external memory and multiplieraccumulator. Data memory/Control Unit is laid out in this paper.

A new ripple analogtodigital converter(ADC) has been developed. It consists of two parallel ADCs and a switching network. The circuit operates on the input signal in two serial steps. First a coarse conversion is made to determine the most significant bits by the first parallel ADC. The results control a switching network to connect the series resistor segment, the analog signal is contained within, to the second parallel ADC. At second step, a fine conversion is made to determine the least signification bits by the second parallel ADC. The circuit requires 2(2
$\frac{N}{2}$ ) comparators, 2(2$\frac{N}{2}$ ) resistors, and 2(2$\frac{N}{2}$ ) switches for Nbit resolution. 
This paper proposes testable design method of Domino CMOS NORNOR Array Logic design method. Previous Domino CMOS PLA method is composed of 2 level NANDNAND Logic. Because NORNOR Logic is realized by a parallel circuit, this method can prevent delay time each level and DNORPLA include testable circuit system that DNORPLA circuit. DNORPLA testable algorithm is realized on Prime (Primos) in Pascal language and DNORPLA circuit is simulated by PSPICE.

This paper proposes the logic simulation for hierarchical logic network with function descriptor base data structure and algorithm on which a macro cell is considered as a logic elements. Function descriptor base data structure is useful when many logic elements of which type is same exist in a network, for it lessens the computer memory size used during the simulation. And the proposed simulation algorithm may improve the simulation speed.

This paper describes KUIC_DRC(Kyungpook national University Intelligent CAD_Design Rule Checker) which verifies VLSI layout. It uses modified linked list data structure. The input form is modifed CIF(Caltech Intermediate Form), called KIF(Kyungpook Intermediate Form). It makes error file, a KIF file. It is written in C language and excuted on MSDOS, in IBM PC/AT.

In the case of VLSI computer system control unit design using PLA, optimal state code assignment algorithm to minimize the PLA area is proposed. An optimal state code assignment algorithm which considers output state and logic minimization simultaneously is proposed, and by means of this, algorithm product term is minimized. Also, by means of this algorithm running time and memory capacitance is decreased as against heuristic state code assignment algorithm which uses matrix calculation and considers the constraint relation only. This algorithm is implemented on VAX 11/750 (UNIX4.3BSD). Through the various test example applied proposed algorithm, the efficiency of this algorithm is shown.

A threestep heuristic algorithm for PLA column folding and row folding of columnfolded PLA is presented, which is significantly faster than the earlier works and provides nearly optimal results. The three steps are i) mincut partition of vertices in the column (or row) intersection graph, ii) determination of products' order using Fiduccia's minnet cut algorithm, and iii) headtail pairing for column folding, while some heuristics are proposed for deciding row folding pairs. The time complexity of this algorithm is O(
$n^{2}$ log n) compared to the O($n^{3}$ )  O($n^{4}$ ) of the earlier works.$^[2][3][9]$ For a test PLA with 23 inputs, 19 outputs and 52 products, the number of column folding pairs obtained using this algorithm is 20 which is optimal, as compared to 17 in a previous work. 
In this paper, a new folding algorithm for design of high density PLA's is proposed. This algorithm generates transform matrix so that programmed region of each row becomes narrow. Based on transform matrix, folding pair is selected. In the process of generating transform matrix, interval degree and dense degree are used to improve folding ratio. This paper also proposes an efficient folding pair graph which has short time of selection folding pair. This proposed algorithm is named PLAFOLII and implemented in C language on VAX11/750 (UNIX) computer. Through various example, efficiency of this algorithm is verified.

In this paper, a flash type A/D converter using Variable Threshold Logic circuit is designed and is layonted by double metal CMOS 2 um design rule. Comparator and register string which is the basic elements of a general flash type A/D converter are substituted by simple comparator circuit which is slightly modified from cmos inverter.

In this paper, a text compression method is proposed which is capable of reducing mean bits per character by word and etymology dictionary. This dictionary consists of 256 words and 512 etymologies with 10 bits length codes. Using this dictionary, a mean rate of 3.44 bits per character is achieved.

This paper proposes a method for analyzing the Korean inflected word in machine translation system. We define the processing rules which are useful of analyzing an irregular conjugation, pesent an parsing algorithm of noun and specifed verb and reduce the space of dictionary by the algorithm.

This paper deals with a semantic properties of Korea noun for semantic process in machine translation. The procedure is carried out as follow; 1) 17,000 words of Korean nouns are collected. 2) Semantic category is classifed into 39 markers. 3) We slow the redundancy of semantic properties and improve the efficiency of dictionary by marking the hierarchical concept structure.

A vowel is based on the recognition of a phoneme. Thus it is necessary for the programming of an algorithm to achieve the speech recognition in that case. In this paper, cepstrum is used for a voicedunvoiced decision and speech parameters are extracted by linear prediction coding. Using these parameters, a speech understanding algorithm has been developed.

This paper describes the implementation of recognition of speaker  dependent Korean spoken continuous digits. The recognition system can be divided into two parts, acoustic  phonetic processor and lexical decoder. Acoustic  phonetic processor calculates the feature vectors from input speech signal and the performs frame labelling and phone labelling. Frame labelling is performed by Bayesian classification method and phone labelling is performed using labelled frame and posteriori probability. The lexical decoder accepts segments (phones) from acoustic  phonetic processor and decodes its lexical structure through phone network which is constructed from phonetic representation of ten digits. The experiment carried out with two sets of 4continuous digits, each set is composed of 35 patterns. An evaluation of the system yielded a pattern accuracy of about 80 percent resulting from a word accuracy of about 95 percent.

This paper describes a fingerprint classification on the basis of feature points(whorl, core) and feature vector and uses a syntactic approach to identify the shape of flow line around the core. Fingerprint image is divided into 8 by 8 subregions and fingerprint region is separated from background. For each subregion of fingerprint region, the dominant ridge direction is obtained to use the slit window quantized in 8 direction and relaxation is performed to correct ridge direction code. Feature points(whorl, core, delta) are found from the ridge direction code. First classification procedure divides the types of fingerprint into 4 class based on whorl and cores. The shape of flow line around the core is obtained by tracing for the fingerprint which has one core or two core and is represented as string. If the string is acceptable by LR(1) parser, feature vector is obtained from feature points(whorl, core, delta) and the shape of flow line around the core. Feature vector is used hierarchically and linearly to classify fingerprint again. The experiment resulted in 97.3 percentages of sucessful classification for 71 fingerprint impressions.

이 논문에서는 필기체 한글 인식에 있어서 입력된 기본자소를 window를 이용한 윤곽선 추적과 삼각분할에 의한 이분점 추출에 의해 각 기본자소가 갖고있는 특징성분을 찾아내고 그 특징성분에 의해 문자의 골격을 추출하여 인식하는 방법을 제안하였다. 윤곽선 추적시 window를 이용함으로 간단한 잡음제거와 추적속도를 증가 시켰으며 삼각분할에 의한 이분점 추출방법을 사용함으로 단순한 윤곽선 추적에 의해 특징성분을 추출하는 방법보다 문자의 특징성분을 정확하게 추출할 수 있다는 장점을 갖는다.

This paper suggests the operation performing the noise elimination of binary image. The image is read by the scanner. And operand is selected according to the size of input image. Through the Dilation and Erosion, elementary vector operation with selected operand, the noise of input image is eliminated.

This paper is concerned with distributed INGRES database system. The main motivation is to provide dynamic reconfiguration and global database consistency. Dynamic reconfiguration will provide proper initial environment to newly generated system. By use of transaction characteristic, database can be maintained in consistent state.

This paper has been aimed to develop M68020 Software Development System on a host computer 3B20S for the TDX2 fully electronic Switching system by transporting M68020 C Cross Compiler. M68020 C Cross Compiler source code which includes assembler, runtime library and optimizer has been analyzed for the installation on the host computer 3B20S system. Moreover, the linkage editor source file has been analyzed and installed on the3B20S to produce the executable file correctly. Through these procedures, the M68020 object codes could be obtained on the 3B20S computer for the multiusing purposes. It has also been confirmed that the M68020 Software Development System on the 3B20S works correctly.

This paper proposes and sketches out a new parallel architecture of transputerbased pyramidal parallel array computer (TPPAC) used to process computationally intensive problems for geometric processing applications such as computer vision, image processing etc. It explores how efficiently the pyramid computer architecture is designed using transputer chips, and poses a new interconnection scheme for TPPAC without using additional transputers.

This paper proposes a microcode generating system which automatically generates the microcode of various target machine by inputing the intermediate language (MDIL) from the machine independent HLMLC (High Level Microprograming Language C) language. The MOP's (Microoperations) which is modeled 7tuples generate to extend MDIL by table driven method with the information of translation table for each target machine. As compaction being considered and the hardware resource of target machine used, the conflicts of hardware elements are removed possibly. This proposed system is implemented with C language and yacc on VAX11/750 (UNIX 4.3 BSD).

Interconnection network is to provide communication among functional modules. The interconnections considered are Generalized Cube networks. Two situations are examined: a memory module is equally likely to be addressed by a processor and a processor has a favorite memory. This paper proposes the effective condition of operation in interconnection network through performance evaluation by simulation.

SPM is MDC Processor for string pattern expressed in directional chain code. In this paper we consider the string pattern matching algorithm (Levenstein Algorithm) whitch is portion of Dynamic Programing, and propose architecture of SPM and simulate it on the RT level to evaluate its architecture. We used the C language as the hardware description language, and developed it on the IBM PC/AT Zenix system V OS environment.

This paper proposes the MinimumDistance Classification(MDC) processor using the Residue Number System(RNS). The proposed MDC Processor in this paper is efficient for realtime pattern clustering application and illustrate satisfiable error rate in application experiments of image segmentation but error rate increase as cluster number do.

A new tridiagonal Toeplitz linear system (TTLS) solver is proposed. The solver decomposes a strictly diagonally dominant TTLS equation into a number of subsystems using a limit convergent of an analytic solution of a continued fraction. Subsystem equations can be solved employing a modified Gaussian elimination method. The solver fully exploits parallelism. Optimized operational environment for the algorithm is discussed.

This paper describes a matching method to solve the problem of occlusion in a two dimensional scene. The technique consist of three steps: generation of hypotheses, clustering of hypotheses by matching probability, updating of hypotheses. Using this algorithm, simulation results have been tested for 20 scenes contained the 80 models, and have obtained 95% of properly correct recognition rate in average.

This paper describes the method of syntacticsemantic pattern recognition and description for two dimensional object which is adjusted or changed in size and its orientation. To avoid the complexity and ambiguity which is arised in the case of syntactic or decisiontheoretic method is used individually, an attributed grammar is introduced which applies computative attributes to pattern primitives, and then uses decisiontheoretic method for attributes and syntactic method for pattern structure. A primitive extraction embedding parsing and grobal rule for classification is also applied for more effective pattern recognition and description.

From the viewpoint of general recognition system, Bspline is introduced for the approximation and recognition of human profile contours. Profile contour is approximated to the cublic Bspline curve by least square fitting so that BSpline nodes nearly correspond with the curvature extrema of the contour. This method is designed for the spline to be good features in recognition, and also showed good approximation compared with the variants of Bspline appraximation.

In the recognition of similar patterns like profile contours of human faces, feature measure plays important role. We extracted effective and general feature by Bspline approximation. The nodes and vertices of the approximated curve are normalized and used as features. Since the features have both local property of curvature extrema and global property by Bspline approximation, they are superior to those of curvature extrema of the profile contour. For the image data of six sets of 56 persons, some of which are illmade, averaged accuracy rate of 97.6 % is obtained in recognizing combinational 333 test samples.

Algorithms that represent given pattern in the form of an ARG (Attributed relational graph) using not only structural relations but also symbolic or numerical attributes, and then recognize that pattern by graph matching process are presented in this paper. Based on definitions of pattern deformational models, algorithms that can find GPECI(Graph preserved error correcting isomorphism). SGECI(subgraph ECI) and DSECI(Double subgraph ECI) are proposed and comparisons among these algorithms are described. To be useful in performig practical tasks, efficient schemes for extraction of ARG representation fron raw image are needed. In this study, given patterns are restricted within objects having distinct skeleton, and then the information which is necessary for recognition and analysis is successfully extracted.

This paper proposes the model of Image Display on the Xenix System V and describes the method of writing device driver which can drive the graphics hardware and image scanner. The interface between user program and device driver in Kernel is also presented.

본 연구에서는 구조적 표현의 화상 처리 알고리즘인 BF(Breadth First) 선형 4진 트리 알고리즘(BFQT 알고리즘)의 압축, 재생부를 하드웨어화 하여 ASIC(Application Specific Integrated Circuit)을 설계한다. ASIC과 IBM PC와의 인터페이스를 명시하며, 새로운 하드웨어 알고리즘을 도입하여 ASIC의 세부구조를 설계한다. 소프트웨어로 수행할 때 보다 제안된 ASIC으로 수행할 때가 압축은 약 21배, 재생은 약 4배 빨라지는 것으로 추정된다.

Ultrasound Doppler Diagnostic System utilizes the Doppler effect for measurement of blood velocity. The sign of the Doppler frequency shift represents blood flow direction. Pulsed Doppler System uses Phase detector and zerocrossing method to produce simultaneous independent audio and velocity signals for forward and reverse blood flow direction in the time domain, had been fabricated. But timedomain analyzing such as audio evaluation and zerocrossing detection for instantaneous and mean frequency measurement doesn't, provide both an accurate and quantitative result. Therefore, it is necessary to adopt frequency domain technique to improve system performance. In this paper, we describe a unit which is composed of Pulsed Doppler System and realtime spectrum analyzer (installed TMS 32010 DSP Chip). This unit shows timedependent spectrum variation and mean velocity of blood Signal.

Ultrasonic Doppler Diagnostic System utilizes the Doppler effect for measurement of blood velocity. The sign of the Doppler frequency shift represents blood flow direction. CW(ContinuousWave) Doppler System uses quadrature detection and phase rotation method to produce simultaneous independent audio and velocity signals for forward and reverse blood flow direction in the timedomain, had been fabricated. But timedomain analyzing such as audio evaluation and zero crossing detection for instantaneous and mean frequnecy measurement do not provide both an accurate and quantitative result. Therefore, it is necessary to adopt frequencydomain technique to improve system performance. In this paper, we describe a unit which is composed of CW Doppler System and realtime spectrum analyzer (installed TMS 32010 DSP Chip). This unit shows timedependent spectrum variation and mean velocity of Blood signal.

To characterize the bioloical tissues, the new methods to measure the frequency dependent attenuation are presented in this paper. In general, ultrasonic phase information was assumed by linear function of the frequency. But the minimum phase function which characterizes the frequency dispersion of tissue was derived in (2). It is very significant to measure the attenuation by using the minimum phase function to characterize the frequence dispersion of tissue. Also, we propose the phase correcting technique to take advantage of the idea that the distortion of amplitude component when the wave propagates through media.

The continued interest in diagnostic applications of ultrasound has led to a need for more detailed knowledge of the ultrasonic properties of biological tissues. In this paper, temperature dependence of ultrasound velocity and attenuation is considered through a fundamental experiment. Acryl and Pig Liver are used as specimen. Data is collected over the temperature range of 2050 C. As a result, we know that ultrasound velocity and attenuation coefficient are uniformly changed in proportion as specimen's temperature.

The Diagnostic applications of Ultrasound are developed in many ways. In this paper, We measure the attenuation coeffient of biological tissue using DSP. This method is useful in tissue characterization with real time. In the future, We expect that this method coupling with the ultrasonic temperature dependence of biological tissue also is applied to hyperthermia.

This paper is a study about the automatic control of wheelchairs. This is realized by joystick, and is simulated by voice signal recognition. The control system by joystick is designed as follows: joystick paddle is connected with a timer and this timer ouput is high only when the joystick is moved. A computer reads the duration of this high state, and ouputs motor control word decided from this value using lookup table. The control system by voice signal is designed as follows: partial autocorrelation coefficients are computed from A/D converted signals and these values are compared with referance patterns. From this, the motor control word is decided on by the neareast neighbor rule.

In this paper, cardiac parameters and relationship between stroke volume and impedance change were explained, and an impedance cardiograph was designed and constructed, and its accuracy was verified by experiment. Useful cardiac parameters such as stroke volume and contractility of cardiac muscle were measured noninvasively. The reproducibility of the instrument was measured to be better(less than 10%) than that of clinical standard method such as thermodilution. Hence impedance cardiography was found to be better technique for monitoring stroke volume and contractility of patients for pre and post operation, and pharmacological studies.

This paper describes a design of hardware system for real time automatic diagnosis of ECG arrhythmia based on pipeline processor consisting of the three microcomputer. ECG data is acquisited by 12 bit A/D converter with hardware QRS triggered detector. Four diagnostic parameters  heart rate, morphology, axis, and ST segment  are used for the classification and the diagnosis of arrhythmia. The functions of the main CPU were distributed and processed with three microcomputers. Therefore the effective data process and the real time process using microcomputer can be obtained. The interconnection structure consisting of two common memory units is designed to decrease the delay time caused by data transfer between processors and by which the delay time can be taken 1 % of one clock period.

This paper is aimed at describing the design and implementation of real time transmission and its processing system for ECG signal. For this purpose the analogue module dealing with 3ch ECG signals, D/A converter module and builtin type MODEM are developed. The result of field tests was good for possible practical uses.

In this paper, we deal with the movements identification of EMG signals by LPC cepstrum coefficients. Movements were identified by extration of characteristics of similar patterns in Euclid distance measurement method for EMG signals generated by voluntary contractions of subject's musculature. As number of coefficients is larger, we obtain the better rate of movements identification. By exact extraction of signals and decision of optimal coefficient, it is expected that these results will apply to prosthesis control in realtime.

Control algorithm for remote transmission processing system for ECG signals is proposed. Software for the system hardware consists of system control algorithm and signal processing algorithm. Since signal processing algorithm is now under developing, this paper describes the details of system control only.

This paper describes an arrhythmia detection algorithm using PP and PR Interval. In order to detect Pwave accurately, an improved 5point derivative method is used. In this paper, the RR, PP and PR interval. and the number of Pwaves per RR Interval are detected for arrhythmia detection. These parameters can be utilized to diagnose in the varied types of AV block, atrial fibrillation, and PVC.

In this paper, the effective ionization coefficients (
${\alpha}{\eta}$ /Po) are measured by the steady state Townsend method in Townsend discharge domain. The effective ionization coefficients are measured in the range 75${\leqq}$ E/Po${\leqq}$ 150(V/Torr. cm) in$SF_{6}$ . The values of the effective ionization coefficients are easily and precisely determined by means of the linearization of current growth equation. The effective ionization coefficients of$SF_{6}$ were agreement with that of Bhalla and Craggs. 
The paper presents the measurement results on the parameters affecting the breakdown mechanism of a large air spacing under switching impulse voltages. Measured parameters are the velocities of leader channels, predischarge currents, electric charges injected into the rodplane air gap and electric field intensities on the plane. For the 3m air gap under switching impulse voltages, the velocities of leader channel have been measured to be of 1cm/
${\mu}s$  5cm/${\mu}s$ , electric field intensity of 2kv/cm, predischarge current of 1.2A  1.6A, the charges injected into the air gap of 11  40${\mu}$ C for 400887kV impulse voltages. 
This paper is to investigate dielectric properties, dielectric breakdown strength, varing of interior organization and gelling point temperature as parameter of postcure conditions of thermosetting epoxy resin.

Thermally stimulated current, cubical expension, capability change were measured by temperature variation. According to the capability change, TSC peak value was increased. We found that the crystal dissolution is 375 K and amorphous state becomes 388K. Charged partical behavior in the dipole and electronic trop were found iomic conduction in the low field and electronic conduction in the high field. Charged particle in the semiconduction storey was aceumulated in the interface by electron injection which can be arise TSC.

Electron mobility which is a basic parameter to represent gas state, is calcurated using equation assumed Lorentz gas. This equation is applied to 30mm 20W fluorescent lamp. In order to obtain electric field strength of positive column, lamp and cathode fall voltage wave form is measuread using indirect voltage measuring methode. To obtain electron density, used the calculation of S.C Peek and D.E Spencer. The result which is applaied 38mm fluorescent lamp at 400mA,
$20^{\circ}C$ is displayed. 
The Pockel's effect is well know as a voltage measurement method. This method is applicable to high voltage measurement equipments. We observe the Faraday rotation of the polarization plane in the currentinduced magnetic field.

The Mechanical strength and the electrical insulation properties of the Epoxy Resin vary with the formation of chemical reactions between Epoxy and curing agent. Therefore, to study their reactions is important for understanding the properties of the cured Epoxy. In this paper, to get the cured Epoxy having good electrical insulation property, it is studied the effects of the surrounding temperature during the cure process on the property of the cured Epoxy. In the experiment, Epoxy is cured with the surrounding temperature varied between 5[
$^{\circ}C$ ] to 50[$^{\circ}C$ ]. As the results, it follows that the higher the temperature is, the faster the crosslink process is. 
This paper covers the insulation coordination of 765 KV transmission line in Korea Electric Power Corporation. The design for the lightning, switching surge, and contamination was conducted to increase the reliability of 765 KV line. The authors have used the result of a limited Westinghouse ANACOM swiching surge study of the Korea Electric Corporation's 765 KV Expansion plan. The study investigated the switching surge overvoltage distribution and maximum switching overvoltage when relosing 765 KV lines. They used the technical data of the suspension insulator for the contamination design, which was available from a foreign insulator manufacture. From this paper, the authors find out that the contamination design dominates to determine the number of insulator and the SOV design dominates the air clearance in the tower. Afterwards the authors will confirm this design by Demonstration Test with Korea Electrotechnology Research Institute.

The formation and properties of porous silicon layer(PSL) formed by anodic reaction in hydrofluoric acid solution have been studied. Many micropores are formed randomly inside of PSL and the anodization is achieved uniformlytoward the thickness direction. Current density, resistivity and HF concentration in Ptype PSL formation are found to play important roles in determining the formation and properties of PSL.

The capacitancevoltage (CV) chracteristics of thin nitrided thermal oxides prepared by rapid termal nitridation(RTN) have been studied. The threshold voltages were calculated using CV measurement and found to vary as the concentration of acceptor and the thickness of oxynitride. When the Si02 films were annealed in NH3 a decrease in the positive oxide charge due to SiN bond was observed. In the case applied frequency is high and low, the high frequency depletion capacitance was higher than that of low frequency, which is indicative of high frequency surface conduction by mobile surface charge.

The CT characteristics of nitridized MOS capacitor have been studied. The generation lifetimes were calculated using CT transient response ans found to vary as sample condition. This is due to the nonuniformity of fast surface state. Also, This experimental curves were different from theoretical curves. The result suggests that the change in material structure (from SiO2 to SiNO) is important in improving minority carrier lifetime.

Titanium dioxide antireflection (AR) Coating, which is deposited on Si substrates using an organotitanium solution by the spinning technique, has been studied. The coated films on Si substrates were subsequently heated to
$450^{\circ}C$ . The thickness and index of refraction of films were varied continuousely from$740{\AA}$ to$1380{\AA}$ and from 1.7 to 2.1 respectively as a function of heat treatment temperature and time. Silicon solar cells ARcoated by the spinning technique showed as much as 31% improvement in conversion efficiency over the uncoated cell. 
LangmuirBlodgett(LB) films of arachic acid and TCNQ(tetracyanoquinodimethane) were prepared in the sample of Al/LB film/Al type where Al are electrodes, and polarization in LB film and dipolar moment of molecules in the films were measured by themally stimulated current(TSC). It is ascertaind in our experiments that dipolar mament of
$C_{12}$ TCNQ molecule was about 1315 debye and the moment was directed from the alkyle chain toward TCNQ. Generation of the open circuit voltage was investigated on the ztype and Hetero structure of LB films. 
This paper was studied on the effects of CaLaNb substitution and Zr/Ti ratio variation to Pb(Zr, Ti)
$O_3$ system on structural, electrical, optical and sound level characteristics in order to develope the piezoelectric and electrooptic ceramic devices. Also the specimens were prepared by the two stage sintering technique. The molecular formular was X($CaO{\cdot}1/4La_{2}O_{3}{\cdot}1/4Nb_{2}O_{5}){\cdot}(1X)Pb(Zr_{Y}Ti_{1Y})O_{3}$ (x=100X, y=100Y), and the variation of x was$6{\sim}12$ , y was 60${\sim}$ 49 and second stage sintering time was 20${\sim}$ 40 hours. The experimental results obtained from this study are as follows : 1. The density was decreased, the grain size was increased according to increase of CaLaNb substitution. 2. The crystal structure was rhombohedral in composition 6/60/40, and the crystal structure was tetragonal and cubic according to increase of CaLaNb substitution. 3. The Ca substitution of PZT system enhanced the sintering property. The Pb site vacancy resulting from the substitution of LaNb increased the dielectrical constant, the piezoelectric charge constant, the dielectric loss and decreased the coercive field. 4. The resistivity of PZT system which has the P type conduction mechanism increased according to substitution of LaNb because of the substituent acting as donor. 5. The PZT ceramics varied from ferroelectric substance according to increase of CaLaNb substituent. The coercive field and saturation remanent polarization decreased, and at last straight line according to increase of LaNb substitution. 6. The amount of CaLaNb substitution to improve the light transmittance of speciment was 10 mol%, the Zr/Ti ratio was 49/51, and the second stage sintering time was 40 hours. 7. According to CaLaNb substitution, the specimens was to be transparent. The 7.5/51/49 specimen was suitable for transparent sound vibrator because it had 58% light transmittance (thick 0.2[mm], wave length 700[mm]) and 48% electromechanical coupling factor. 
The objective of the present study was to manufacture the piezoelectric composites of Lead Zirconate Titanate (PZT)  polymer for applications such as ultrasonic medical diagnosis and hydrophone. The PZT rodpolymer composites have been prepared with 4.5 to 57 volume percent PZT using 1.5 mm rod, so the connectivity pattern of the two phase is 13 type. The electromechanical coupling factors of radial mode (
$k_{p}$ ) and thickness mode ($k_{t}$ ) were nearly independent of the volume percent PZT, which were 0.3 and 0.65, respectively. The acoustic impedance of the piezoelectric composites was measured by the resonance technique in the frequency range 50 KHz1.5 MHz, which was in the range of 3.8  60 Mrayl. 
The frequency response characteristics of a bimorph type PZT piezoelectric transducer was investigated. In this study, function generator which generates short burst signal, plane reflection plate and oscilloscope were used to measure the characteristics of piezoceramic ultrasonic transducer. The resonant frequency of a bimorph type piezoceramic transducer which is acquired by using ToneBurst Method had good agreement with the measured results from spectrum analyzer.

This paper is to investigate dielectric properties and new fabrication method of GlassCloth/Epoxy composite materials. According to the results of this paper, gelling point temperature is affected significantly on the ambient temperature. And Tan
${\delta}$ characteristics of GlassColth/Epoxy composite materials is also affected significantly on the state of interior void of glass cloth and interface coupling between epoxy resin and glass cloth. 
This study is to investigate the adhesive strength of composite material's interface on the experimental methode of tree growth in the material. The results are as follows 1) The irradiations of ultrasonic energy cause the mechanical vibration in the polymer composite materials of fluid state, so then bring about physical dispersion and heat for inorganic materials, being supposed to produce chemical interlinking reaction, decreasing of voids between filler and matrix. 2) As the intensity of ultrasonic energy and its irradiated time are larger, the tree inception and breakdown voltages increase and the tree growing is slower. so we obtain that the interface adhesive force can be strengthened by the irradiation of ultrasonic energy.

Thermal Electet of the Zinc Oxide varistor has been studied in the temperature range of 130
${\sim}200[^{\circ}C]$ and the electric field of 6[kV/m]. It appears that there are four peaks of thermally stimulated current;${\alpha},\;{\beta},\;{\gamma}$ and${\delta}$ spectra appearing at the temperature range of 160, 130, 20 and$30[^{\circ}C]$ , respectively. It seems that the origins of${\alpha},\;{\beta},\;{\gamma}$ and${\delta}$ peaks are associated with the depolarization of donor ions in the depletion layer, the detrapping of trapped electron in the surface, the detrapping of trapped electron in the donor level and the detrapping of trapped electron between grain and intergranular, respectively. 
The electrical properties of ZnO varistors fabricated by the second firing method were investigated. The nonlinear coefficient of ZnO varistor fabricated by this method is similar to that of commercial ZnO varistor. But the breakdown voltage is higher than that of commercial ZnO varistor. These results are attributed to grain boundary diffusion of
$Bi_{2}O_{3}$ by second firing. 
Eddy Current Sensor to Measure the
${\gamma}{\alpha}$ / Transformation Ratio of Metal is Designed and Implemented. Experimental Results Regarding to Impedance Variation due to${\gamma}{\alpha}$ Transformation Ratio of Metal Under Room Temperature are Reported. It Was Found That the Impedance of Metal is Increased With Increasing${\gamma}{\alpha}$ Ratio. 
This paper considered temperaturesensitive characteristics on the basis of curie tempeature and quenching method in the process of manufacturing MnCuZn Fersite. The results are as follow. Curie tempeature drops according as the content of CuO and ZnO increases. It also decreases according as sintering temperature increases when the content of ZnO in fixed. Curie temperature drops more in quenching than in slow cooling and activation energy diminishes were too. On the basis of curie tempeature, activation energy is greater in paramagnetic region than in ferrimagnetic region. As its voltagecrrent characteristics is similar to that semiconductors, the temperaturesensitive ferrite is expeated to be appied in the area of power electronics.

In this paper, we introduced high quality earth leakage breaker(ELB). We improved performances of the ELB by strengthening the environmental stress. We suggested that it have the electromagnetic noise filter, finally high sensitive ZCT for asymmetric fault conditions (including do component).

We study both theoretically and experimentally the motion of fluid driven by electrothermally ablated gas pressure in a cavity with a single exhauster. A possibility of jet propulsion engine for steamers is also discussed as an application of the fluid motion.

For determination of blackout area and restoration area by an expert system for fault section estimation and power system restoration using information from circuit breakers, it is necessary that the recognition of system isolation state and a method of finding the change of system isolation state by the state transition of breakers in isolated system. This paper presents a method of resolving the above problem by rulebased approach.

In electromagnetic transient analysis studies, complex transmission system should be modelled in detail. But in large system, modelling of complete system is a big burden in many sides such as computation time, modelling efforts. It is needed to model only a concerned part of the system in detail. The remaining parts of the system have to be reduced to equivalents. This study suggests a method of generating a simple network equivalents which has equivalent frequency charastics in wide frequency range.

The simplified simulation technique for the best generation mix is developed and the studied results are described. The best generation mix over study time from the economic point of view can easily constructed by this technique.

An expert system is developed to solve the voltage problems occurring in electric transmission power system. It is based on knewledge engineering technique satisfies the performance criteria such as minimizing the number of operation of control device and quantity of reactive power. Also, it uses bestfirst search technique from the experts. Control devices used in this paper include shunt capaciter/reactor, transformer tap changer, generator output voltage and a generator is used after the availability of other two devices are checked.

Two methods for probabilistic maintenance scheduling are developed and compared ; one with operation and suppliedshortage cost and other with risk level of LOLP. Based on the real economic power dispatch, quadratic optimal maintenance conditions are obtained and simple amtrix equations are suggested for solutions. Both methods are compared in a sample system of 26,000 [MW] peak and 32,000 [MW] generation capacity.

This paper discusses one overall reactive power installation pattern that will satisfy system performance constraints for base case. The cost of installation is minimized, Existing controllers are fully utilized prior to any addition of new installation. Installation pattern for the required new shunt compensation is comprised of discrete, standard size components.

This paper presents a optimal Var allocation algorithm for minimizing transmission line losses and improving voltage profile in a given system. In this paper, nodal input data is considered as Gaussian distribution with their mean value and their variance. A Stocastic Linear programming technique based on chance constrained method is applied, to solve the var allocation problem with probabilistic constraint. The test result in 6Bus Model system showes that the voltage distribution of load buses is improved and the power loss is more reduced than before var allocation.

In the operation of a power system, the security of the system has acquired significant importance to supply electric power of better quality. The State Estimator, a part of security functions, provides a complete real time solution estimate of the steadystate conditions of the power system for use by the Real Time Network Analysis functions. This paper briefly introduces the Fast Decoupled Weighted Least Square State Estimator which is adopted in the KEPCO EMS with features of ManMachine Interface.

AESOPS computer program is designed to find those eigenvalues of a linear dynamic system model which most strongly characterize oscillations between generator rotors. The low frequency oscillation actually occurred in October, 1986 in Korean Electric Power Corporation (KEPCO) System. 28 oscillation modes are identified through computer runs and two of them are calculated unstable in the case of the start of a pumpedstorage unit. This program is considered to complement the conventional PSS/E stability program for the stability study of KEPCO system.

In this paper, we constructed the distribution database using computer graphics which is supported by a microcomputer system. Using this micro computer system, we established the retrievement technique of the transformer and customer information. We showed visually the displaying technique for the separation and restoration of the distribution system fault zone.

This paper describes a fast security analysis techique for voltage security assessment. The new method identifies the location of buses with potential voltage problems and thereby defines a voltagesensitive subnetwork for contingency screening. The efficieny of this method is derived from the use of a voltage subnetwork to drastically reduce the number of bus voltages to be solved; and subsequently from the use of compensation techniques and sparsevectors methods for screening. Results demonstrating the effectiveness of the method on the IEEE14 bus model system are presented.

The results of the first demonstrative project of photovoltaic power generation system for the remote islands (Hahwado, Jeonnam Province) were reported. The optimum design of a 25 KWp PV system and a 5 KWp diesel backup system were discussed.

Recently, with the expansion of electronic equipment and computer facilities, good quality of services is required. This paper describes how to improve distribution system voltage by the Tap position seletion of P. Tr, by the decision of the optimal sending voltage in substation and we also developed three phase load flow program in distribution system. Besides we have developed the digital recording voltmeter that will be utilized with personal computer system. This meter will be useful, economical and powerful.

This paper is concerned with linear system containing time delay in model reference adaptive control. An algorithm is developed using Lyapunov second method and properties of strictly positive real function.

Up to now, using adaptive control method, Identification deals with system whose entire state variables and prameters are accessible for measurement. In practical situations, all the state variables cannot be measured and it is impossible to directly apply since the parameters of the system are unknown. Therefore, in this paper, using only inputoutput data, such a model of the system is not available since the parameters of the system are unknown. this leads to the concept of an adptive observer in which both the parameters and the state variable of the system are identified simultaniously. Lyapunov's direct method and KalmanYakubovich (KY) lemma are employed to ensure the stability of this schemes. The feature is that the signal and adaptive gain which is generated from filter is imposed upon feedback vector and then state variables and the unknown parameters can be identified. To show the usefulness of the proposed schemes, computer simulation result of unknown secondorder system shows the effectiveness of the proposed schems.

For a generalized minimum variance controller algorithm the weighting polynomials are are calculated in a way to assign the closed loop poles of the system and to specify the controller gain at a frequency. As a result the oscillations in the control signal may be reduced without changing the deterministic behaviour of the system.

In recent, Kalman filter technique has been much used as one of technique for tracking of the moving target. But some problem are still remained to be resolved. For example, when Kalman filter technique is applied to nonlinear system, the technique is nonoptimal estimator. Therefore, extended Kalman filter is proposed to reduce modeling error for nonlinear system. In this study, an extended Kalman filter in Cartesian coordinates is described for moving target, when the radar sensor measures range, azimuth and elevation angle in polar coordinates. And an approximate gain computation algorithm is proposed. In this approach, Kalman gains are computed for three uncoupled filter and multiplied by a Jacobian transformation determined from the measured target position and orientation.

This paper proposed the new control law which is defined by instantaneous output and time delayed output. To analyze the system with time delayed output, the way which transforms output time delayed systems into instantaneous output systems is presented. The output responses were more improved by the new control law then that of the instantaneous output control law. The algorithm for simulation and a numerical example are presented.

Indirect adaptive pole assignment PID controllers for unknown systems with time varying delay, is proposed. Unknown system parameters are estimated by recursive least square method, and time varying delay is estimated using indirect predictors. Since the order of parameter vectors didn't increase, the computational burden is not largely increased in spite of using indirect adaptive control method with time varying delay estimation. Computer simulation is performed to illustrate the efficiency of the proposed method.

The effects of the errors due to incorrect a priori informations on the noise model as well as the system model in the continuous state model based optimal FIR filter is considered. When the optimal filter is perturbed, the error covariance is derived. From this equation, the performance of the state model based optimal FIR filter is analyzed for the given modeling error. Also the state model based optimal FIR filter is compared to the standard Kalman filter by an example.

An attractive approach to speed of induction motors is to achieve full linearization via appropriate feedback. However, the prior results toward this direction are based on full feedback. In practice, rotor fluxes are not directly measurable but can be estimated using observers. We propose a nonlinear feedback controller with an observer. As t
${\rightarrow}{\infty}$ , the closedloop system with our controller becomes as if it were a linearly decoupled system. We provide the stability analysis of our control method. Simulation and experimental results are also included to demonstrate the practical significance of our results. 
In this paper, the implementation of NIU (Network Interface Unit) for Token Passing Bus network is studied. The network is based on the IEEE 802.4 standard, which is the basis of MAP. IBMPC is chosen as host station. The structure of the network and NIU is investigated and the implementation is illustrated. The logical ring constructed following the standard. Transmission and reception of data are tested. The experimental results are referred.

This paper demonstrates an AI application for representing operator's heuristic knowledges in implementing process control. AI controller constitutes servo and regulatory controller, respectively. The knowledge base for servo controller was designed to obtain the process output with respect to the desired setpoint in short period. The regulatory controller was to provide smooth output near the setpoint. The AI controllers was implemented in TurboPROLOG on IBMPC. For the parallel processing, TurboC was linked to calculate the process output. The result shows that this AI controller can be well suited not only for minimum phase but also for nonminimum phase. This controller may be used as a backup controller for rather extensive expert system.

The problem of minimization of energy, consumed by the robot manipulator, is important, notably for larger manipulators, higher working speeds, and loads. Obviously, the stated problem requires the application of optimal control theory, which is being successfully applied for linear system and certain classes of nonlinear systems. However, the application of optimal control theory(in energy or time optimization) leads to substantial practical difficulties, so that significant simplifications are usually performed, either in model complexity or by neglecting the existing constraints. In this paper the problem of obtaining such an optimization method. which would take into account the complete system dynamics and all the constraints is considered. The only method found to be suitable for such a complex optimization should be based on dynamic programming.

To assure a successful completion of an assigned task without interruption, the hand of mechnical manipulator often travels along a preplanned path. This paper presents a method of obtaining a time sechdule of velocities and acceleration along the path, under the constraints. Because of the involvement of a linear performance index and a largr number of nonlinear constraints, the "method of approximate programming (AMP)" is applied. To overcome the false solutions AMP is modified. To reduce the computing time, a "direct approximate programming algorithm (DAPA)" is developed.

In this paper, collision avoidance of two robots working in a common workspace is described. The workspace for a two robot system is defined and classified. JCM (Joint Constraint Map) is designed which shows the region of joint values that must be avoided for collision free motion, and application of the JCM is studied.

The equations of motion for linearly elastic bodies undergoing large displacement motion are derived. This produces a set of equations which are efficient to numerically integrate. The equations for the elastic bodies are formulated and simplified to provide as much efficiency as possible in their numerical solution. A futher efficiency is obtained through the use of floating reference frame. The equation are presented in two forms for numerical integration. 1) Explicit numerical integration 2) Implicit numerical integration. In this paper, there was used the numerical integration. The implicit numerical integration is extended to solved second order equation, futher reducing the numerical effort required. The formulation given is seen to be occulate and is expected to be efficient for many types of problems.

A jigsaw puzzle matching technique is proposed. Specifically, the geometric patterns of the puzzle pieces are firstly extracted using a boundary tracking algorithm at low resolution. And then, features of the extracted pieces to describe jigsaw puzzle pieces such as angles and distances between corner points, and convexity or concavity of a corner point are obtained from some corner points implying discontinuity of curvature of puzzle pieces' boundary. Finally, a boundary matching algorithm without a priori information of matched puzzle is proposed.

This paper describes an automatic visual inspection system for fabric defects based on pattern recognition techniques. The inspection for fabric defects can be separated into three sequences of operations which are the detection of fabric defects[1], the classification of figures of fabric defects, and the classification of fabric defects. Comparing projections of defectdetected images with the predefined complex, the classification accuracy of figures of fabric defects was found to be 95.3 percent. Employing the Bayes classifier using cluster shade in SGLDM and variance in decorrelation method as features, the classification accuracy of regional figure defects was found to be 82.4 percent. Finally, some experimental results for line and dispersed figures of fabric defects are included.

This paper describes the development of a prototype expert system, which is designed to analyze loom production efficiency by employing a loom monitoring system. The problem field of the prototype expert system is the loom failure analysis which has a great effect on prouction efficiency, and the characteristic graph in which the loom failure causes are classified and organized, is used in the analysis. The knowledge for characteristic graph is organized, and an effective inference engine is developed. The developed system is tested and modified using hypothetical loom failure causes, and the feasibility of the expert system application to extend the functions of the loom monitoring system is proved.

This paper deals with automatic system of acoustic noise detection for realization of factory automation. The existing inspection process of failure products has mostly been executed in hand by richexperienced workers. It is difficult to accomplish effectively or systematically the failure test of products owing to the diversality of illconditions. But the problem about it must be solved in viewpoint of cost down and factory automation in addition to the reliability of products. The necessity of automatic inspection system to inspect automatically undesirable acoustic noise of products which is one of the kinds of failure is suggested, and the procedure to develope it and the function of each system components are explained briefly.

In this study, we designed a model of NC controller with IBMPC as a host CPU and intelligent servo controller for 2 axes that can be expaned to 4 axes. OS software was developed with C language at the base of mode selection technic for 9 NC operating modes. Servo controller design was based on the application of interpolator IC(3701) and position controller IC(3702) that permits low cost and high performance. For connection of two systems, parallel I/O communication was implemented. Finally, auto interpolation program test was executed for linear and circular paths resulting 1 LSB accuracy.

This paper summarizes economic considerations for building automation. While the market for energy management systems(EMS) has been exploding, there are many indications that the pendulum of hight expectations may have swung too far. This is a very important reason for energy management system in both existing and new buildings. Because facilities continue to expand in size and their operations grow more complex, the use of automated systems to control and monitor both programmable and nonprogrammed function becomes an economic necessity. The following is offered as engineerings of economic consideration for building automation systems. Not only get a complete return on investment but increase profits in a few years through savings incurred in labor, utilities fuel and extended equipment life.

In this paper, a control system of the PUMA 560 robot manipulator using a PC (Personal Computer) is presented. The hardware of the designed control system is composed of IBMPC/AT, interface board, selection board, interrupt generating circuit, and the servo control unit of the PUMA controller. A robot control library is developed using assembly and C language, and is composed of several lowlevel functions and arm interface routines, world model routines, arm kinematics routines, and motion command service routines. Using the designed library, joint interpolated motion and Cartesian interpolated motion of the PUMA robot manipulator are realized. In the future, our system is expected to be a very helpful basis and a useful tool for developing various control algorithms of robot manipulator using sensory information.

The new controller which conststs of IBMPC/AT, SBC, servo module and motor driver has been developed for KIMMBOT II robot. The servo module which controls two axes at the same time also has beendeveloped. This module inputs relative number of pulse and feedrate via multbus, after then outputs analog command voltage with reference to feedback pulses from the encoder of a motor. This system has a great flexibility. The reduction of parts gives more reliability.

RMFC (Resolved Motion Force Control) is the method to control the Cartesian force and position using FCC (Force Convergent Control) instead of the complicated dynamic equations of the manipulator. The gain parameters of the controller are adjusted through many trial and errors. In this paper PDoptimal control method is introduced to give optimal gain parameters which minimize the difference between actural acceleration and desired acceleration. To show the validitiesn of the proposed method computer simulations are performed for the twolink manipulator.

The basic concept of this control scheme is that of an ideal massless manipulator carrying a load. The purpose of this paper is to converge the observed forces and moments at the mass center to the desired applied forces and moments. And the corresponding control algorithm is proposed.

A robust control method for robots in presented. In this method, a phaselag controller is used for reducing the effect of the unknown payload without the measurement of joint accelerations and torque/force. Simulation results for the lower 3 joints of PUMA 560 show considerable reduction of position errors due to the unknown payload, compared to the computedtorque method.