Electromagnetic wave scattering from the two-dimensional scatterer was calculated by the Boundary Element Method (BEM). For the circular cylindrical scatterer, the BEM solutions agreed very well with the analytic solutions. The rectangular dielectric cylinder was also treated in the case of the lossy scatterer and the lossy medium.

The problem of electromagnetic scattering from arbitrarily shaped lossy dielectric bodies of revolution is studied. Two coupled vector integral equations are formulated from the potential formulas, the equivalence principle and boundary condition. The unknown surface currents (both electric and magnetic) can be found by the moment method, Galerkins's procedure. After the surface currents are determined, the scattering pattern can be evaluated.

New measurement method of coupled transmission line characteristics is described. This method presents precision calculated values of even- and odd-mode impedances and effective dielectric constants of mictostrip parallel coupled lines from the scalar quantities obtained by transmission coefficients at two different resonance frequencies. Measured impedances and effective dielctric constants are good agreement with predicted values.

This paper concerns methods for ${\mu}$-wave imaging. The image reconstruction of an object by range-doppler preceding using the X-Band Linear-FM signal is presented from tile simulated data. The high degree of range resolution is achived using large signal band width and cross-range resolution is obtained by doppler processing.

Theoretical calculations are presented for analying lateral modes of stripe geometry lasers. The solution technique affords a matching between the fields of the active layer and those of the surrounding passive layer. The fields are written as a liner combination of Hermite-Gaussian function. Therefore fields have been described with a single H - G function. The lowest-order mode spreading is calculated and related to the gain distribution.

Recently Hopfield proposed a model for content-addressable memory, which has been shown to be capable of storing information in a distributed fashion and determining the nearest-neighbor. Its application is, however, inherently limited to the case that the number of l's in each stored vector is nearly the same as the number of O's in that vector. If not the case, the model has high probability of failure in finding the nearest-neighbor. In this work, a modification of the Hopfield's model, which works well irrespective of the number of l's (or O's) in each stored vector, is suggested.

There are various holographic recording materials, such as Dichromated Gelatin, Silver Halide, Thermoplastic, Photoresist. In this paper especially, we used Photoresist to make the phase holographic grating. Deep-groove diffractive grating formed in relatively thin holographic recording material is to express high diffraction efficiency. Phase holographic grating recorded In photoresist can be used very practical because it has the high diffraction efficiency, and it is possible to make a replication easily. So, it has the merit in recording the optical holographic grating than any other materials.

A new method optical short pulse generation using diode lasers under the negative electro-optic feedback are proposed. The self-pulsing is induced by feedback itself and the pulse is formed by the interaction of feeback signal and the dynamics of the diode laser. The simulated pulse width is on the order of picosecond with several GIIz repetition rates.

In this paper, we develope a system which can perform signal processing for bar-code laser scanner. This system is composed of optical detector and preprocessor. The former detects the diffused light and converts it into TTL lebel output. The latter discriminator valid data from various raw data and transmits data to micro-processor. The preprocessor consists of edge transition detector, latch signal generator, module counter, register array, adder array, and buffer memory control circuit etc..

We demonstrate the optical bistability and multistability using light emitting devices, a photodetector, and transistors under the positive electro optic feedback. Its operating principle and high speed operation Scheme are also described.

A parallel-plate waveguide with a slit in its upper plate is analysed. The magnetic current induced in the slit is obtained by making use of the conventional moment method. From knowledge of the magnetic current, quantities of interest such as reflection coefficient and transmission coefficient are computed.

The scattering characteristics are analysed for the dielectric cylinder with arbitrary cross section partially covered by thin conductors. The integro differential equations consistant with boundary conditions of conductor and dielectric boundaries are derived by the equivalence principle. They are transformed into matrix equations by moment method. The circular dielectric cylinder covered by conductors at the upper and bottom side of the cylinder is chosen for the numerical example. Current distributions on conductors and scattering cross section by the cylinder are computed for incident wave perpendicular to the conductor plane.

Electromagnetic waves from the satellite may be utilized to direct the tracking antenna toward the satellite. We design and fabricate the feed appropriate to the monopulse tracking technique which derive angle-error information on the basis of a single pulse. The feed consists of five cavity-backed turnstile elements mounted on a common ground plane. The turnstile dipole are connected to a set of five quadrature hybrids which convert the dual linear polarization into dual circular polarization. The five feed outputs are then processed in the monopulse comparator which is constructed in microstrip for compactness.

In this paper, feature vectors composed of 6 features of Fourier spectrum of 2-D image at each projection angle and 7 features of invariant moments are defined. The feature are extracted by optical Fourier transformer and Radon transformer. After extracting the feature, the input pattern is recognized using the squared Mahalanobis distance.

Phase conjugate wavefront generation via real-time holography in degenerate four-wave mixing experiments in photo-refractive crystal $Bi_{12}$ Si $O_{20}$ is presented. Depending on the respective values of applied field $E_o$ and fringe spacing, drift or diffusion of the photecarriers dominates the space-charge buildup and affects the Phasa- Conjugate Wavefront intensity. The experimental results for wavefront reflectivity as a function of the intensity ratio of the writing beams agree with the theoretical predictions.

An acousto optic correlator system for performing real-time correlation of two signals is constructed. It uses AOD is described. The statistical analysis for correlator's output SNR is presented and experimental procedure for implemention of this technique are result shown. Since the correlation intergration is performed in time rather than space, the SNR of this system can be greatly improved.

A granite optical table with a vibration isolator system is designed and constrvcted. The table top is made with granite and the vibration isolation system is constrvcted using a pneumatic isolator. The performance of the table is evalvated by the vibration measurement.

The problem considered is that of calculating the angles of arrival of a multimoded signal at an antenna array from simultaneous measurements of the amplitude and the phase of the signal in each element of the array. The method of wavefront analysis 15 applied to a 4-element direction finder of the Adcock type. A simulated problem is constructed with the aid of suitable computer programs developed here.

A solution which combines ray and aperture integration(AI) techniques is presented for the problem of electromagnetic plane wave scattering by an open-ended, perfectly-conducting, semi-infinite parallel plate waveguide with a thin, uniform layer of lossy or absorbing material on its inner walls, and with a simple planar termination inside. Numerical results are given for the fields outside the waveguide.

A method for synthesizing low-pass elliptic filters in a microstrip configuration is presented. The realization consists of the cascade connection of proper rectangular elements, each one corresponding to four reactive elements of the lumped-constant prototype. This allows an effictive control of parasitics and unwanted reactances. Which results in the possibility of realizing higher order filters with cutoff frequencies UP to X-band. Fifth and seventh order filters were fabricated on allumina substrates.

By cascading the step discontinuities on planar dielectric waveguides with a ferrite layer through two uniform planar dielectric waveguides, nonreciprocal scaterring characteristics of planar dielectric waveguide gratings are analyzed. The calculated results at 35GHz shows that the isolation mechanism takes place near band stop frequency range.

We investigate the feedback linearizability of nonlinear discrete-time system s of a specific form, $x_k=G_{u_k}oF(x_k)$ where F is a diffeomorphism and [$G_{u_k}$] forms an one parameter group of diffeomorphisms. This structure represents a class of systems which are state equivalent to linear ones and approximates the sampled data model of a continuous-time system. It is also considered a relationship between linearizability and discretization.

In this paper, "Revised two-level costate prediction method" is developed to optimize the quadratic performance of a class of nonlinear dynamic systems. To show the merit, of this algorithm, the proposed algorithm is compared With "The new prediction method" and "Two-level costate prediction method". Advantages of this algorithm are illustrated by applying it to three examples, turbine generator system, fermentation Process, power control system in nuclear reactor.

In this paper, the stability margins of LQ regulators for the systems with delays in both state and control are analyzed and represented explicitly in terms of system parameters when the systems are open-loop stable. And, the LQG/LTR method is considered as a robust control design method. The results in this paper generalize the well-known ones for ordinary systems.

An adaptive control algorithm for a plant with unmodelled dynamics is proposed. The upper bounds of the output due to the unmodelled dynamics and measurement noise is assumed to be known. Linear programming is used in estimating the bounds of plant parameters. Projection type algorithm is used in estimating the plant parameter with these bounds. This algorithm is nearly the same as those proposed by Kreisselmeier or Middleton except that the bounds are computed by linear programming. The stability of the proposed algorithm Can be proved in nearly the same way as that of Middleton. Simulation results show that the proposed algorithm gives better parameter convergence and smaller overshoot in the plant output than the algorithm without computing the bounds of plant parameters by linear programming.

In this paper, the pole assignment control algorithm with a discrete integral action is discussed in that it reduces the impact of an external step disturbance upon the output. The sensitivity of output is viewed according to changes of delay-time which this algorithm is applied to a continuous plant with.

The characteristic of a control system is apt to be changed by load condition. This paper getting more optimal environment and the algorithm for parameter by itself In PID control system with microprocessor.

An analysis and design of large-scale linear multivariable systems often requires to be block triangularized form for good sensitivity of the systems when their poles and zeros are varied. But the decomposition algorithms presented up to now need a procedure of permutation, rescaling and a solution of nonlinear algebraic equations, which are usually burden. To avoid these problem, in this paper we develop a newly alternative block triangular decomposition algorithm which used the generalized matrix sign function on the Z-plane. Also, the decomposition algorithm demonstrated using the fifth order linear model of a distillation tower system.

This paper describes a new method for control of large-scale system by sliding mode. The concepts of control to large-scale system on the basis of VSS(Variable Structure System) control theory are used to decompose a large control problem into a two-level algorithm such that each subsystem is stabilized with local discontinuous controllers and higher level corrective control is designed to take into account the effect of interaction among the subsystems. In this paper, we show that each subsystem is controlled with repect to local continuous and higher level corrective control. This algorithm can be easily applied to multi-variable control system and obtained a continuous control in comparison With variable structure control systems. Two numerical examples are discussed as illustrations.

A new method of model reduction, based on the differentiation of polynomials, is Introduced by Gutman(1982). Without the reciprocal transformation, and the differentiation of the numerator and denominator polynomials of the transfer function, used by Gutman, the lower-order system can be obtained by the permutation of the order of numerator and denominator and the number of differentiation.

This paper presents the choice of appropriate sets of input variables for large-scale linear multivariable systems. It is shown that the selection of a good set of input variables for control may become important when both strong and weak input variables are available. The transmission of information from the inputs to the outputs is investigated and appropriate scaling procedures to derive a scaled input matrix are proposed. Singular value decomposition methods facilitate the quantification of the systems excitation stemming from the various input variables, and thus the selection of an appropriately strong and orthogonal set of input variables. The need for and the implementation and benefits of reducing the number of input variables are illustrated by a large-scale steam generator model of a real process.

To improve the reliability of boiler controller of a power plant, FTCS(Fault Tolerant Control System) is proposed. We studied to implement a Multi-processor scheme for FTCS. This paper presents the total system to experiment the performance of FTCS and the Multi-processor scheme implemented.

This paper presents a control scheme to stabilize a two-axes microwave antenna system which is located on a ship and is subject to roll, pitch and yaw(RPY) motions. The scheme first computes the compensated angular positions of the antenna for RPY disturbances through coordinate transformations, and then use a servo con roller of PID type for each axis to track the compensated position command with a good transient behavior.

As a means of improving the reliability of the analog type controller for the thermal power plant, an efficient method is proposed, which is to place the hardware redundancy, i.e. a back-up controller with fault detecting capability. FTCS is implemented by using multi-processors and it is experimentally verified that the back-up controller takes over the role of the original controller, controlling the faulty loop.

A graphic display and operator console system is developed for monitoring the operation of power plant. It has multiprocessor structure using VME bus and common memory. The graphic monitoring system is applied to fault tolerant control system for enhancing reliability of boiler analog controller. As a result, it displays all the operating date as color graphic images with 14 pages. Moreover, it can transfer the operator commands to the other micro-processors through common memory.

In this paper, the performance of Mini MAP architecture in realtime application is analyzed by analytic methods. The approximate equations for token rotation time and waiting time are obtained and using these equations network. performance is evaluated. According to the results obtained, Mini MAP architecture has good realtime performance and can be used to many real time applications.

This paper was aimed to, using a new data structure, develop a set of algorithms to execute the output function of Digital System. These functions were represented as directed, acyclic graphs. by applying many restrictions on vertices on graph, the efficient manipulation of boolean function was accomplished. The results were as follows; 1. A canonical representation of a boolean function was created by the reduction algorithm. 2. The operation of two functions was accomplished using t he apply algorithm, according to the binary operator. 3. The arguments having 1 as the value nf function were enumerated using the satisfy algorithm. 4. Composing TTL 74181 4-bit ALU and 74182 look-ahead carry generator, the ALU having 4-bit and 16-bit as word size was implemented.

This paper addresses the identification of various constraints in time-varying obstacle avoidance for mechanical manipulators. The manipulator constraints include the smoothness constraint and torque constraint, while the environmental constraints include a motion priority, a traveling time constraint, a path constraint, and a collision constraint. The inherent difficulties in combining these constraints are discussed with a suggestion for the purpose of time-varying obstacle avoidance.

Recently various control algorithms for industrial manipulators have been proposed. However, computation time, modelling error, and torque type controller design have prevented real-time implementation. As the result, most of performance evaluations of control algorithms have been carried out only by computer simulations. In this paper, we explore real-time implementation to show the feasibility and effectiveness of such algorithms. Experimental results indicate that computed torque method and learning control algorithms can be effectively applied to control industrial manipulators.

The force/torque sensor for robot is developed. The compliance algorithm for peg-in-hole insertion task using the forec/troque sensor is developed. The system consists of an IBM PC, robot, force/torque sensor, strain meter, A/D board, and interface board. The IBM PC functions as a main processor and the robot controller as a slave processor. The sensor is constructed to measure $T_x$, $T_y$, $F_z$ which are necessary to precisely execute a peg-in-hole insertion task by SCARA type assembly robot. The outputs of sensor are analyzed. On the basis of the analysis, compliance algorithm for peg-in-hole insertion task is developed. Some comments concerning the development of wrist force/torque sensor and compliance algorithm are given.

High performance requirements such as high speed operation. accuracy and versatility have led to the consideration of structural flexibility in robot arms. The purpose of this study is to investigate the interrelationships between the robot structural flexibility and a linear controller for the rigid body motion. This paper employs an assumed modes method to model both the rigid and flexible motion of the robot arm. The simulation results illustrate the differences between leadscrew driven and unconstrainted axes of the robot.

In this paper, a servo controller for a robot manipulator is developed. It consists of a conventional MRAS controller for each joint together with a feedback control to Dynamics compensation. To generale the control torque effectively a currant drive method is adopted. The speed of the actuator is measured by using a tachometer and the position of the link is measured by using a potentiometer. To show the effectiveness of the proposed control algorithm the proposed controller is tested in real time.

Algorithms to detect in real-time both position and orientation of rectangular type electronic components are developed for industrial vision. In order to conduct detection in real-time, parallel processing algorithm of image date which uses several control processor is proposed. Image processing area is divided into several regions which can be processed by each cpu. As a result, processing time is improved when two control processors are used and real-time pattern recognition of not-well-aligned components is accomplished.

In an industrial vision system, a coding technique for binary image is essential to extract useful informations. To reduce the processing time, a hardware implementation of the chain coding algorithm is attemped. For that purpose, the chain coding algorithm is modified so that it is more suitable for a hardware implementation. A hardwired chain coder is also developed and tested with developed vision system. The result shows that the processing time is greatly reduced and that the developed vision system is maybe feasible for real-time applications.

In this paper, the design method, design techniques and structure of a language for a SCARA type industrial robot, are presented. The proposed new language is modular and expandable using the C programming language and the 8086 assembly language. It is composed of monitor mode, editor mode, execution mode, I/O mode and teach mode. The developed language is implemented on the robot controller to verify its performance.

In this paper, a programming system for SCARA-type robots is designed, consisting of robot language, computational facilities and programming tools for handling interconnection environments. In designing the robot language, CLRC(C Library for Robot Control) is introduced, using the general-purpose language 'C' as base programming language. Also the motion primitives for Continuous Path control as well as Point-To-Point motion arc included. By means of frame and homogeneous transformations the system is capable of applying the SCARA-type robot efficiently and easily for any given task.

In this paper an intelligent supervisory programmable controller for automatic assembly machine is developed. This is achieved by adding sequence control hardware with input-interrupts to supervisory real time language and also by incorporating an automatic planning method which uses a predicate logic model and an action model. The designed intelligent supervisory programmable controller is applied to Die Bonding Machine and is to found to work well.

In this paper, the design and implementation of a multiprocessor based Die Bonder Machine for the semiconductor will be described. This the partial research result, that is, the 1st year portion of the project to be performed for a period of two years from June, 1986 to May, 1988. The mechanical system consists of the following three subsystems : (i) transfer head unit, (ii) die feeding XY-table unit, and (iii) plunge up unit. The overall control system is designed to be essentially a master-slave type in which each slave is functionally fixed in view of software and also the time shared common bus structure with hardwired bus arbitration scheme is utilized, the control system consists of the following three subsystems each of which employs a 16 bits microprocessor MC 68000 : (i) die bonder processor controller, (ii) visual recognition/inspection and display system, (iii) the servo control system. It is reported that the proposed control system were applied to Working Sample and tested in real system, and the results are successful as a working sample phase.

The purpose of this paper is to develope the parallel computation algorithm that enables it to minimize the completion tine of computation execution of the entire subtasks, under the constraints of the series-parallel precedence relation in each subtask. The developed algorithm was applied to the control of a robot manipulator functioned by multi-CPU's and to obtain the minimum time schedule so that real time control may be achieved. The completion time of computation execution was minimized by applying "Variable" Branch and Bound algorithm which was developed In this paper in determining the optimum ordered schedule for each CPU.

A microprocessor-based incremental servo system with a variable structure is presented. New system properties are obtained by composing a desired trajectory from the parts of trajectories of different structures. It is implemented with Z-80A microprocessor to control a dc motor, and fast response without overshoot and good steady-state accuracy can be easily obtained.

This paper deals with a development of auto mobile watch system which can control direct auto mobile state or position. Main controller consists of the Z-80 computer, sensor and high frequency receiver. Mobile loading consists of the Intel 8751H one-chip and ultrasonic transmitter, having able to detect all auto mobile state. And, the search data is sent the main controller by the high frequency transmitter and ultrasonic transducer.

In this paper an efficient algorithm to estimate the volume and surface area and the reconstruction algorithm for 3-dimensional graphics are presented. The graph theory is used to estimate the optimal quantitative factors. To improve the computing efficiency, the algorithm to get proper contour points is performed by applying several tolerances. The search and the given arc cost is limited according to the change of curvature of the cross-sectional contour. For mathematical model, these algorithms for volume estimation based on polyhedral approximation are applied to the selected optimal surface. The results show that the values of the volume and surface area for tolerances 1.0005, 1.001 and 1.002 approximate to values for tolerances 1.000 resulting in small errors. The reconstructed three-dimensional images are sparse and consist of larger triangular tiles between two cross sections as tolerance is increasing.

The energy gaps of undoped and Cr doped GaAs and $Al_{0-3}Ga_{0-7}As$ was measured nondestructively using surface acoustic wave (SAW) technique. The result is good agreement with other reported result. This technique is inexpensive and simple one for the measurement of mole fraction of $Al_xGa_{1-x}As$.

CMOS operational amplifier is most useful building bloch in analog circuit. This paper represents the analysis and design method of CMOS OP AMP to use general purpose such as the A/D and D/A converter, PCM encoder and decoder etc. The required specifications is obtained by changing W/L ration of CMOS devices. The design procedure must be iterative in as much as it is almost impossible to relate all specifications simultaneously. This is performanced with IBM-PC XT by using SPICE(SIMULATION PROGRAM WITH INTEGRATED CIRCUIT EMPHASIS)program.

A layout design concept which suppress the substrate current generated during the switching transients of an CMOS inverter is presented. The amount of hot carriers and the peak value of substrate current can be reduced by changing the device geometry ratio of driver and load device of an CMOS inverter.

A $32{\times}33$ MOS-type area image sensor has been fabricated. The blooming current is reduced to 1/14 by forming +p photocell in P-well instead of a simple p-type substrate. A shallow n+ junction is made to improve the sensitivity of photodiode on short wavelength. Bootstrapping circuit technique is applied to obtain high speed dynamic shift register. The shift register operates at up to 10MHz for 7V clock.

The miniaturization of Bird's Beak generated at the field oxidation has been studied by adding polysilicon layer between the silicon nitride and pad oxide stack, which is the basic structure of Conventional LOCOS. The size and shape of Bird's Beak were intensively observed by SEM, and also the electrical characteristics of Bird's Beak-reduced LOCOS structure were compared with those of Conventional LOCOS. The length of Bird's Beak was reached up to $0.20-0.28{\mu}m$, while about to $0.50-0.53{\mu}m$ in conventional LOCOS, resulting in 60% reduction.

Planarization in multilevel metallization is very important to smooth out topographic undulations by conductors, dielectrics, contacts, and vias. One of methods for planarizing interdielectrics, such as the etchback process of the double layer composed of the photoresist on the interdielectric low temperature oxide was introduced. The step heights of interdielectrics before and after etch-back process was measured by Scanning Electron Microscope, and the degree of planarization was analyzed, comparing the differences of the step heights. In this experiment, the degree of planarization was controlled up to about 0.9.

A $2{\mu}m$ CMOS P-well double metal technology has been developed. Phosphorus deep implantation and drive-in diffusion steps were utilized to prevent the low voltage bulk punch through in the short channel, 1.6[${\mu}m$] Leff, PMOS device. Double metal process with the rules of 5[${\mu}m$] 1st metal pitch and 7[${\mu}m$] 2nd metal pitch was successfully implemented by using VLTO, low temperature oxide, as on intermetal dielectric.

A new LOCOS (Local Oxidation of Silicon) process using a thin nitride film directly deposited on the silicon substrate by LPCVD has been developed in order to reduce the bird's beak length. SEM studies showed that nitride thickness of 50nm can decrease the bird's beak length down to 0.2um with 450nm field oxide. No crystalline defects are observed around the bird's beak after the Wright etch. A 30% improvement in current density was obtained when this new method was applied to MOS transistors (W/L*2.9/20.4) compared to conventional LOCOS process (bird's beak length=0.7um). Other various electrical parameters improved by this new simple LOCOS process are reported in this paper.

Thermal reoxidation was carried out to eliminate hole traps at the surface of Si wafer. As the result, the good surface state of wafer was obtained and hole traps were eliminate at the inversion layer. For the evaluation of reoxidation effects. MOS diode was fabricated and its C-Y curve was plotted.

The thermal oxidation recipe has been optimized for very thin (12 nm) capacitor oxide for Mega bit DRAM. The time dependent dielectric breakdown characteristics show that the breakdown voltage and time to breakdown are very high and uniform, indication that our oxide is defect free and suitable for DRAM capacitor dielectric. To our knowledge this is the best oxide quality obtained up tp now around 10 nm.

$PbTiO_3$ thin film is prepared by rf sputtering method to implement the pyroelectric infrared detector at room temperature. Annealing of $PbTiO_3$ thin film is done from $400^{\circ}C$ to $550^{\circ}C$ each for 2 hours in furnace. The spectral response to recrystallization process of $PbTiO_3$ thin film is measured by IR photospectro meter. Pyroelectric detector Modeling is studied for implementing device using electrical equivalent circuit model. It is found that $PbTiO_3$ thin film has two IR absorption band within $1000-400\;cm^{-1}$ (10um-25um) and it's spectral response is improved as annealing temperature increase. As a result of pyroelectric detector modeling, we find the possibility of implementing optimum device structure.

Formation of the titanium silicides was performed by the furnace annealing. Ti-silicide was formed by reacting Ti films with singlecrystalline silicon in vacuum or nitrogen ambient in the temperature range $500{\sim}900^{\circ}C$. The Ti-Si interaction in such films was investigated by using X-ray diffraction, and sheet resistance measurements. It was found that the dorminant crystal phase of silicide formed during annealing at $600{\sim}700^{\circ}C$ was TiSi, and $TiSi_2$ phase is associated with a very low sheet resistance(<$2{\Omega}/{\Box}$).

A two-dimensional numerical simulation program is presented which enables the concentrion of impurity in Silicon to be caculated. The numerical algorithm used is based on the boundary element menthod.

With the fabrication of Al-$Al_2O_3$-n(p) type Si devices, the analysis and measurement of various characteristics, this study presented the electric physical property theory for the charge distribution of MAS device $Al_2O_3$ films, and inquired out the devices available. In order to study them, Al-(450A)$Al_2O_3$-n(p) type Si was the main objects in the study. They were examined through carrier injection, C-V curves of devices on time, ${\Delta}V_{FB}$-t curves, I-V curves and $Al_2O_3$ film's breakdown characteristics.

This paper describes a model to calculate the equilibrium electron density of MODFET at the interface that takes into account the simultaneous shallow and deep level in the Al-GaAs layer. In the present study we have made an investigation of the interface electron density with different values of the AlGaAs doping density and spacer layer thickness, considering simultaneously two doner levels. In this case, the ratio of the shallow to the deep donor concentraction is considered. From the comparison with early experimental results we could find the deep level and that the deep donor concentration is about 50% with the Al mole fraction X ${\sim}0.3$, activation energy Edx=65meV, temperature $77^{\circ}K$ and spacer thickness range $50A{\sim}100A$. Also we have investigated the effect of the temperature. As temperature increase, at critical mole fraction X the nature of the donor concentration changes from $\Gamma$ to L and X.

The hot-carrier induced degradation in very short-channel MOSFET was studied systematically. Under the traditional DC stress conditions, the threshold voltage shift (${\Delta}Vt$) and the transconductance degradation (${\Delta}Gm$/(Gmo-${\Delta}Gm$)) were confirmed to depend exponentially on the stress time and the dependency between the two parameters was proved to be linear. And the degradation due to the DC stress across gate and drain was studied. As the AC dynamic process is more realistic in actual device operation, the effects of dynamic stresses were studied.

Low temperature phosphorus-nitride CVD was newly developed for a high quality gate insulator on InP substrate. This film showed the Poole-Frenkel type conduction in high electric field with resistivity higher than $1{\times}10^{14}$ ohm-cm near the electric field of $1{\times}10^7\;volt/cm$. The C-V hysteresis width was very small as 0.17 volt. The density of interface trap states was $2{\times}10^{11}cm^{-2}ev^{-1}$ below the conduction band edge of InP substrate. Effective electron mobility was about $1200-1500\;cm^2/Vsec$ and showed the instability of PN-InP MISFETs drain current reduced less than 10 percent for the period $0.5-10^3sec$.

${\gamma}-Fe_2O_3$ thick film city gas sensors have been fabricated by using screen printing method. The sensitivity to $C_4H_{10}$ and, $CH_4$ was about 90% and 65% respectively. The devices heated in air at $400^{\circ}C$ for 1 hour exhibited the highest sensitivity (-90%) to $C_4H_{10}$ at the operating temperature of $300-350^{\circ}C$. And they had good selectivity for $C_4H_{10}$ in comparistion with other gases (CO, $H_2$, $C_3H_8$ etc).

The fabrication of a micro mass flow sensor on a silicon chip by means of micro-machining technology is described on this paper. The operation of micro mass flow sensor is based on the heat transfer from a heated chip to a fluid. The temperature differences on the chip is a measure for the flow velocity in a plane parallel with the chip surface. An anisotropic etching technigue was used for the formation of the V-type groove in this fabrication. The micro mass flow sensor is made up of two main parts ; A thin glass plate embodying the connecting parts and mass flow sensor parts in silicon chip. This sensor have a very small size and a neglible dead space. Micro mass flow sensor can fabricate on silicon chip by micro machining technology too.

For longterm conservation of the $Si_3N_4$ gate pH-ISFET, characteristics of ISFET surface treatment with HF and surface coating with paraffin wax were investigated.

Urea-Sensor was fabricated by immobilizing urease on ISFET's gate using BSA(bovine serum albumin) and glutaraldehyde, and its characteristics were examined. This sensor showed approximately linear characteristic in the urea concentration range of $3{\times}10^{-5}-10^{-9}$ (g/ml). Fast response time was obtained and minute amounts of expensive enzyme were used in comparison to general electrode type biosensors.

Theoretical analysis was performed to understand the sensing mechanism of pH-ISFET by using the site binding model. The calculated pH-${\varphi}$ relationship showed good fittings with the experimental results.

Photoconductive properties of B-doped hydrogenated amorphous silicon thin films prepared by the RF glow discharge decomposition of $SiH_4-B_2H_6$ gas mixtures have been investigated. Experimental results showed that the B-doped hydrogenated amorphous silicon thin films have high photosensitivity and good optical absorption. Therefore these thin films can be used for the photoconductive layer of the vidicon target.

The photoconductor is made of evaporated amorphous selenium as the base material, doped with arsenic and tellurium to prevent crystallization and to increase the red sensitivity of the amorphous selenium. The four-layered photoconductor of Se-As-Te has good photosensitivity(r=0.9) and high dark resistivity($P_d=10^{12}{\Omega}{\cdot}cm$). therefore, this four-layered photoconductor can be used for the target of color image pick-up tube.

The structure of vacuum evaporated PbO films is investigated. Also, the Photosensitivity and dark resistivity are measured. The dominant structure and orientation of these films were red (tetragonal) form, and <110> and <101> direction under- the suitable deposition conditions. And the crystallite size of them was about $2um{\times}0.2um$. These PbO films have good photosensitivity(r=0.9) and high dark resistivity. (${\rho}=10^{14}{\Omega}{\cdot}cm$)

We report that the optimum substrate temperature to fabricate a-Si:H $n^+-p-p^+$ cell decreases with increasing the boron concentration in the Player. The results can be explained as the dependence of substrate temperature for the relaxation of silicon atoms and the bonded hydrogen concentration in the player.

Hydrogenated amorphous silicon (a-Si:H) films have been deposited, for thye first time, by a remote plasma chemical vapor deposition. The hydrogen radical play a important role to control the deposition rate, The bonded hydrogen content to silicon is independent of hydrogen partial pressure in the plasma. Optical gap of deposited a-Si:H lies between 1.7eV and 1.8eV and all samples have sharp absorption edge. B-doped a-Si:H films by a RPECVD has a high doping efficiency compared with plasma CVD. The Fermi level of 100ppm B-doped film lies at 0.5eV above valence band edge.

A laser interferometry of very high spatial resolution was built to study the solid phase epitaxial regrowth of furnace-annealed silicon wafers. As boron concentration increases up to $1.0E20/cm^3$, regrowth rate increases, but pre-exponential term and activation energy are decreased.

The recrystallization of polysilicon layer deposited on Si was attemped by means of C02 laser annealing. The polysilicon layer was defined in small island patterns ($50{\mu}m{\times}200{\mu}m$) by means of photolithography prior to the annealings. After the annealing an increase of grain size up to about 50um was obtained.

P+ ion implanted poly-Si flim with doses of 10E13-10E16 ions/$cm^2$ were annealed by CO2 laser and their sheet resistances were measured and compared with thoses of furnace annealed samples. In case of lightly doped samples, the measured sheet resistance of laser annealed samples were lower several orders of magnitude than those of furnace annealed samples. The origin of this reduction of sheet resistances is supposed to be the increase of the grain size to the extent of certain critical value.

THE ELECTRICAL CHARACTERISTICS Of RAPID THERMAL OXIDES AND NITRIDED OXIDES HAVE BEEN INVESTIGATED. R.T.OXIDE FILMS HAVE BEEN PREPARED BY ONLY R.T. OXIDATION OR R.T.OXIDATION AND SUBSEQUENT R.T.ANNEAL. NITRIDED OXIDE FILMS HAVE BEEN PREPARED BY R.T.OXIDATION AND SUBSEQUENT R.T.NITRIDATION.AND CONVENTIONAL OXIDES ALSO HAVE BEEN PREPARED TO COMPARE WITH R.T.P OXIDES. R.T.ANNEALED OXIDES SHOW EXCELLENT BREAKDOWN FIELD. LEAKAGE CURRENT AND TDDB CHARACTERISTICS. ALSO, CAPACITANCE Of R.T NITRIDED OXIDES ARE SUPERIOR BY 10% TO CONVENTIONAL OXIDES, BUT TDDB CHARACTERISTIC ARE POORER THAN OXIDE FILMS.

We have studied 1iquid phase regrowth of phosphorus ion implanted silicon films on insulator (SOI) by rapid thermal annealing (RTA) method. Many twin boundaries were observed on the regrown silicon layer and mobility of the layer was increased from $14\;cm^2/v.sec$ to $38\;cm^2/v.sec$ after annealing at $1150^{\circ}C$ for 15 sec.

TO INCREASE THE CELL CAPACITANCE Of SMALL GEOMETRY DRAMS. HIGH DIELECTRIC MATERIAL HAS BEEN USED RECENTLY. THE PURPOSE Of THIS WORK IS TO INVESTIGATE THE STRUCTURAL AND ELECTRICAL CHARACTERISTICS Of SiO2/Si3N4/SiO2 STACKED FILM UTILIZING HIGH DIELECTRIC MATERIAL Si3N4(${\epsilon}=7.5$). IN RESULT, THE DIELECTRIC CONSTANT Of STACKED FILM IS 4.0 - 5.0 AND CAPACITANCE AND BREAKDOWN FIELD WERE MORE INCREASED THAN THOSE Of SiO2 FILM.

In the standard system of low voltage-oriented ZnO varistor,a small amount of $Sb_2O_3$ was added to improve the nonlinear exponent and then to find the variation of breakdown characteristics, 0.1m/o-SiO and 0.1m/o-$TiO_2$, respectively,were added We considered relationship between the breakdown voltage of systems and the microstructure. We found that the system containing 0.1m/o-$Sb_2O_3$ showed very high nonlinear exponent. And we found that SiO enhanced breakdown voltage and $TiO_2$ lowered it.

To enhance the breakdown properties of low voltage-oriented ZnO varistor, the samples were fabricated with the amounts of si-oxides and the sintering conditions. And then, to lower the breakdown voltage the $TiO_2$-added samples were fabricated. We investigated the nonlinear exponent, the nonlinear resistance and the V-I characteristics of samples. And we discussed with microstructures by use of SEM and the position of Si by EDS. Si-oxides, especially, largely enhanced the nonlinear exponent. In this case optimum sintering condition was $1200-1250^{\circ}C$-1hr and $TiO_2$ addition lowered the breakdown voltage.

The deposition and reflow properties or BPSG film deposited by APCVD was characterized by variation of each process parameter. As deposition temperature is increased higher, deposition rate is decreased. Maximum deposition rate of BPSG film is obtained in higher 02/Hyride ratio than CVD Oxide or PSG. BPSG film shows stable dielectric properties and we obtained good planarization effect at lower reflow temperature in case of BPSG film than PSG film.

In order to determine the standard composition of ZnO varistor with higher nonlinearity, various contents of $MnO_2$, $Co_2O_3$ were added to ZnO-1.0m/o $Bi_2O_3$ system. Also, samples that contained small amount of Sb, Si-oxides in standard composition determined before were fabricated. As a result, the standard composition of higher nonlinearity-oriented ZnO varistor was shown as ZnO-1.0 m/o $Bi_2O_3$-1.0m/o $MnO_2$-1.0m/o $Co_2O_3$ and $Sb_2O_3$ largely enhanced nonlinear exponent and nonlinear resistance, hut SiO largely enhanced nonlinear exponent only.

A probabilistic production costing model based on the economic load dispatch has been developed. Objective function is composed of fuel cost which is a function of generation output and the failure cost. Coefficients of the failure cost is determined from the known equivalent generation cost. The model is compared with other existing methodolgies and the excellent results are obtained.

In this paper, the circuit analysis during the retrace period of E.H.T. generator is performed by means of the simplified equivalent circuit. The E.H.T. voltage and the beam current are computed by using Newton-Raphson method and the computed results are compared with those of experiment.

The induction motor is represented by nonlinear equations whose parameters are changing with respect to the slip-frequency, temperature, etc. The slip-frequency is effected by unknown load which is difficult to estimate on-line. Astable vector controller is designed with direct MRAC to improve the quality of the transient response. The unknown load is considered in this speed controller design, and tested by simulation. Also a flux controller is designed and tested to reduce the audible noise in this paper.

A volocity and an efficiency of an induction motor are lower than those of a DC motor. But in the control, the utility of an AC motor is more enlarged with the development of inverter. But unfortunately, in using the induction motor, it include a lot of harmonics because of switching of semi-conductor devices(SCR or Tr) in the inverter. In PCR, as the change of a firing angle $\alpha$, harmonics are changed. Hence, in this thesis analyze the harmonics and provide the region of it's application when PWM and 6-step applied to inverter. Of course, frequencies of inverter changed. Especially, in PWM a carrier ratio is changed.

In this paper, a new Harmonic Torque Factor(HTF) that represents torque ripple of induction motor is proposed. Trapezoidal PWM is analyzed to torque ripple by HTF and than we obtain optimum triangular factor at some carrier ratio. Six-step, sinusoidal PWM and Trapezoidal PWM are compared and evaluated for choosing a good control strategy of PWM inverter. It turns out that six-step inverter is better than sinusoidal PWM and Trapezoidal PWM when their carrier ratio is 9, and Trapezoidal PWM is superior to sinusoidal PWM. At induction motor is driven by PWM inverter, torque ripples are investigated by computer simulation using d-q two axis model. It is proved that a new Harmonic Torque Factor ix valid.

Double-diffused metal oxide power semiconductor field effect transistors are used extensively in recent years in various circuit applications. The temperature variation of the drain current at a fixed bais shows both positive and negative resistance characteristics depending on the gate threhold voltage and gate-to source bias voltage. In this study, the decision method of the internal temperature measurement by $V_{GS}$ and $V_{DS}$ are presented.

In phase controlled rectifier(PCR), harmonica are changed according to the variation of a firing angle. These harmonics are supplied with the input of the inverter. And then, inverter output comes about the harmonics combined with the switching frequency of the inverter. Hence the efficiency of the induction motor ia decreased by the harmonic of the inverter output. In this paper, it analyzed about an effect of these harmonics ia analyzed by a computer simulation. The total harmonic distortion (THD) in the case of PCR containing the ripple was considerably larger than THO of the DC source. Therefore, it was proved that the firing angle variation of PCR had to be limited.

An optimization problem minimizing n given time-weighted performance index for discrete-time linear multi-input systems is investigated for the prespecified closed-loop eigenvalues. Necessary conditions for an optimality of the controller that satisfies the specified closed-loop eigenvalues are derived. A computational algorithm solving the optimal constant feedback gain is presented and a numerical example is given to show the effect of a time-weighted performance index on the transient responses.

A new current control scheme is suggest.ed which is suitable for ac servo inverter drives. The scheme uses a simple block diagram to produce the reference stator voltage vector, and the vector nearest to it is chosen for switching. With the same arrangment three kinds of operation modes are possible : (1) constant rate sampling, (2) constant current deviation, (3) adaptive current deviation. In mode (1) current deviation after one period is minimized, while in mode (2) and (3) intervals between switchings are maximized.

교류전동기의 속도조절과 전력절감에 이용되는 인버터는 그 응용분야에 따라 요구되는 출력특성이 다르게 된다. 본 연구에서는 전동기-발전기 셋트와 반환용 인버터를 이용하여 인버터가 응용되는 환경을 임으로 구현할 수 있는 부하장치를 구성하고, 이때 소모되는 전력을 전원에 반환하여 전력소모를 줄었다.

The miniturization of a DC-DC converter circuit in connection with the stability is investigated in this paper. As both the capacitance of the smoothing capacitor and the inductance of the reactor are reduced by raising the switching frequency, it is known that the stability of the buck converter declines with the switching frequency but the buck-boost converter has a nearly uniform stability. Furthermore, that in the frequency region above a certain switching frequency the buck-boost converter is suitable for the miniturization of circuit is cleared.