In this paper, we report the research activities of the technical committee of Electric Vehicle (EV). Recently, environmental pollution has become a world-wide problem. As a practical step to solve this problem, it is sincerely hoped that EVs will become widely used. Here in, we wish to describe an Electric Vehicle research facility with the objectives of developing motor propulsion technology, energy storage technology, control system, battery charging systems, and etc.

요즘엔 신경회로망에대한 소개나 전망등이 신문이나 잡지에도 종종 오르내리지만 불과 몇해전만해도 대학에서 신경회로망에 관한 세미나공고가 붙게되면 공학이나 컴퓨터를 전공하는 사람들조차도 이것은 의학, 또는 신경생리학에 관련된 세미나일것이라 여기는 일이 종종 있었다. 신경회로망의 개발에 대한 시도는 디지탈 컴퓨터와 거의 때를 같이 하지만 세계적으로는 80년대 후반 이래로 그 효용성이 부각되기 시작하였으며 수많은 과학자들이 오늘도 이 새로운 분야의 가능성에 심혈을 기울이고 있다. 그러면 이처럼 전문가 뿐만아니라 일반인들에게도 호기심을 불러일으키고 있는 신경망이란 무엇인가? 이 강좌에서는 그 특성과 구조와 동작원리에 대하여 살펴보기로 하자.

한국의 HAN 인공지능(人工知能)컴퓨터과제나 일본의 NIPT나 성사를 가름할 수 있는 기술 중의 하나가 컴퓨터에 의한 음성인식(音聲認識)의 성공여부이다. 그러나 자동음성인식은 화자독립(話者獨立), 연속음성(連續音聲) 무제한(無制限) 어휘(語彙) 처리라는 세가지 난관을 아직 극복하고 있다. 현재 DTW나 HMM 시스팀은 계속 개선되고있으나 근본적으로 한계가 있다고 보인다. 이와같은 이유로 신경망을 이용한 음성인식연구가 급속히 확산되고 있다. 이와 같은 추세에 따라 본 심포지움에서는 신경망을 이용한 음성인식에 대해 소개한다.

Through the last 20 years' study, it is a well-known fact that symbolic approach has limitations in generating a new concept from given concepts. Hence, neural networks having a role of associative memory based on dynamical activation of neurons attract AI scientists' attention. In this paper, recent trials for combining neural networks and Artificial Intelligent systems are systematically reviewed and a prototype ENEDB(Experimental Neuro Expert DataBase) system built on HP9000/300 workstation is introduced to show the possibility of using the trials for real applications.

Protection and control systems play a prominent part in avoiding power delivery interruptions and help to get a fast and secure restoration when a failure occurs. In order to meet the higher functional requirements on modern power system, protection speed, selectivity, sensitivity, dependability, and security are essential to ensure reliability. These functions on be satisfied by taking advantage of microprocessor and communication technologies, and digital protection relays (systems) have been developed and applied to real power system enhancing reliability and saving money. It is necessary to have a tool to analyze the functions and algorithms of digital relays for installing them to power system. The purpose of this study is to develop a digital relay simulation program to estimate digital relay performances during system faults. Components of digital protective relay including analog filter, sampling unit, digital filter, and relay logic are modeled in this program.

Due to the fault signal containing high and/or slowly decaying dc off-set component, saturation of current transformers cause errors in reproducing secondary current and therefore, secondary current is distorted, which cause relay to maloperate. In this paper, a new method which determines whether current transfomer is saturated is proposed. By designing digital filter of current transfomer, although secondary is distorted, exact secondary current can be reproduced.

Power system fault transient signals are highly distorted due to the presence of high frequency components in the voltage and current signals and an exponentially decaying dc-offset component in the current signals. Modern protective relays have to make reliable fast decisions about the nature of a fault in the presence of such transients. To use a dc-offset removing filter makes relay algorithms much fast and reliable for detecting a fault. In this paper, several dc-offset removing filters are described, and characteristics of them are compared.

This paper presents the real time digital differential protection scheme for transformer. Inrush and Internal fault conditions are distinguished by the relative magnitudes of fundamental and second harmonic components which are extracted from differential currents. The algorithm is simulated and implemented using a prototype relay which is made up of Intel 80286 CPU and Motorola DSP-56001. The testing data of inrush and Internal fault signals which are sampled at a rate of 12 times a cycle are obtained by EMTP, and tested In real time using simulator which downloaded those data.

This paper presents the new protective device which is developed specifically for use in oil filled distribution transformers. This device is mounted Inside of the transformer and connected between the secondary winding and the secondary bushings. It provides both fault and overload protection to the transformer, and visual warning of the existence of uneconomical loading conditions. It Is available in thermal/magnetic operating device depending upon fault duty required.

With regard to the supply of electric power in better quality, the problem of lightning is becoming more and more important in the operation of an electric power system. In this article a description is given of the new lightning position and tracking system which will be installed in KEPCO. The system is based on the principle of time-of-arrival (TOA) rather than the conventional direction-finding (DF) method.

During normal operation of 100% FP Reactor power(TBN/GEN output:690MWe), several times of acceleration phenomena has been generated on the turbine generator-of Wolsong #1 NPP. It was concluded that the acceleration occured following big sudden drop of the terminal voltage of main generator due to AVR potentiometer fault. The cause of turbine acceleration is reviewed with a several records and demonstrated by computer simulation, also presents a countermeasure of its trouble.

The constructions of LLS System is proceeded in cooperation with Korean Meteorological Administration & Korea Electrotechnology Research Institute in Korea. The system is composed of 1 position Analyzer & 10 Direction Finders, and can covers the full area of Korea. We show the configuration of this network and display of data. Finally, we refer to the future research in related with this system in Korea.

The load model representation needs the load composition rate which means the portion of several typical load groups. This paper proposes the refined load composition rate estimation considering the the uncertainty of limit values and an energy portion of a load group.

This paper describes the need of the Database which is designed built with data lot a highly Integrated system for power system planning and operation. The study also, deals plans for and designs of thr highly integrated systems in order to optimize coordination of security and economy of power system planning and operations, as well as to improve the efficiency of data handling

This paper presents an approach for thermal Unit Commitment by Lagrangian Relaxation with fuzzy technique. A proposed algorithm makes it possible to execute optimal decision making between Generation Cost and Load Demand with membership function. In order to test the validity of the proposed method, we applied to Mid-westerm utility system which has 20 thermal units. So, the usefulness of this method is verified.

A planner have to consider many uncertain factors in best generation mix problem. To solve those problems, fuzzy theory is applied in this paper. Using fuzzy measure and Sugeno's fuzzy integral, the planner can choose the best alternative. By applying the proposed algorithm to KEPCO data, the validity of the proposed method is verified.

In the planning stage of bulk power transmission system, reliability evaluation has been a major issue in the sense of suppling reliable power at the minimum cost. MEXICO program, based on Monte-Carlo simulation method, is known as a valuable tool for reliability evaluation. This paper presents the empirical study results when using MEXICO program for IEEE-RTS, IEEE-MRTS and KEPCO-233 system.

Optimal generation dispatch algorithm is developed using L.P and load dispatch method. The algorithm is composed with three steps. Fist phase, analytical algorithm is adopted to solve to minimize quadratic cost functions which fits well for planning purpose. Second phase uses L.P method for obtain economic redispatch that satisfies line constraints. When there is no solution whitch satisfy line constraints load shedding algorithm solves the problem.

The new approach using fuzzy linear programming with fuzzy multi-criterion is proposed for the best generation mix of a power system. A chracteristic feature of the presented approach is that not only cost but also reliability for goal function can be taken into account by using fuzzy multi-criterion and so more realistic solution can be obtained. The effectiveness of the proposed approach is demonstrated by the best generation mix problem of KEPCO-system size model which contains nuclear, coal, LNG, oil and pump-generator hydro plant in multi-years.

This paper Introduces the concept of the generation cost of generating utilities and its calculation methods. Also, the economic evaluation method using generation cost which is called as a screening curve method will be presented along with the benefits and disadvantages of this concept. Next, the least-cost electric utility planning techniques which is used very widely in many countries will be Introduced In comparison with screening curve method. In this aspects, the optimal dynamic mix can be determined as a result. By comparing these two concepts, we will get the concrete concept why the economic evaluation method using generation cost can not be used for the future generation expansion planning.

Recently, increment of power demand is conspicuous, the necessity of demand side management is increased. here, we presented development of economic power demand controller again, we investigated the validity of peak demand control for customers and utilities.

This paper is discribed to the application of PID controller for the frequency control of power systems. The proposed optimal PID controller is designed to obtain gain Parameters of P,I,D's coefficients according to optimization technique using the relatively ingeneous simplex method. The PID controller which is applied to one area, tie-line two area with(or without) exciter is showed good performance rather than that of optimal Pl controllor.

This paper presents an optimal Fuzzy Control Technique to control the load frequency control of multiarea power system with a given stepwise load disturbance. The related simulation results show that the optimized fuzzy control technique are more effective than the conventional control technique (TBC, Optimal Control and etc) for reduction of load frequency deviation in transient and stedy-state, and for minimization of settling time.

Political situations of Korean Peninsula has been changing dramatically in recent years. Interconnected power system operation between South and North Korea should be taken into considerations. This paper discusses several expected operational problems in interconnected systems through literature study. Typical references in automatic generation control area are listed and recommended to read through.

Compared to other sources of electricity such as Thermal and Nuclear power plant, dispersed storage and generation system (DSG) is environmentally clean, quiet and efficient, thus this source is expected to be introduced In urban area by the utilities and customers. If a great number of DSG will be applied to electric power system in the rear future (around 2000, in Korea), these will give a great influence on the existing power system. In other words, the interconnection to electric power system of these source may bring many problems such as system operation, protection coordination, and service quality related with voltages (110${\pm}$6V, 220${\pm}$13V), harmonics (5%) and pour factor (90% over). So, this analysis of the interconnection to power system of DSG is required

Many different industries use Supervisory Control and Data Acqisition/Energy Management Systems (Scada/EMS) to guide a wide range of operations and processes. This paper provides an overview of the functions of Scada/EMS and the fundamentals of operation of Scada/EMS. The paper concludes with the current trends toward open systems, distributed architecture, improved man-machine interface(MMl), advanced applications, artificial intelligence(AI), distribution automation, smarter remote terminal units(RTUs)and expended system scope.

A new algorithm for self-tuning digital controller is proposed. The system to be controlled is identified on line in auto-regressive-moving-average(ARMA) form via recursive least mean square method. The control law is obtained from the minimization of an objective function. The proposed objective function is similar to that of Generalized Minimum Variance(GMV) method but modified to lessen the overshoot and to avoid numerical divergence problem. This algorithm is applied to the power system stabilization and the comparison of the proposed method with a conventional power system stabilizer(PSS) is presented.

This paper represents how to reduce the computer time in small signal stability analysis by selecting the dominant oscillation modes with frequency of 0.5 to 1.2 Hz using the clustering technique. Clustering technique links the buses which are expected to be similar with zero-impedance lines and the voltage variations of these buses are regarded to be identical. The computer time was reduced remarkably with this technique and the effect of clustering will be powerful in the analysis of large-scale power systems.

A Generator Control Systems data is complcated by the existence of a wide variety of Control system types and by the fact that the values of key parameters may depend upon the generator types and operating conditions. This paper decribes the typical value determination methods for generator control systems using response ratio test and step response test. It can be a practical way to provide and verify data for real power system analysis.

The best index should show the margins to the operating limit at the present operating point without much efforts. Such voltage collapse proximity index has been proposed in this study. Differently from the normal procedure in which every bus load is increased at the same propotion, bus load increase toward the direction of T/L loss increase has been applied in this study. Sample studies show the usefulness and the practical applicability in voltage stability analysis.

A simple and practial equivalent technique has been presented in this study. Arbitrary number of bus system can be aggregated to the equivalent one-bus system, equivalents are totel real, and reactive bus powers and T/L losses viewing form boundary buses. Variables to be determined are line impedance between boundary buses and the equivalent bus, voltage and angle of that bus. IEEE 14 bus system is used to demonstrate the proposed method. even in contingency cases very good results have been obtained

Transient stability analysis of Korea Electric power Corporation(KEPCO) system is conducted by time simulation method, and the method is robust and reliable. But, time simulation consumes enormous computing resources and engineering time, and it does not provide a measure of the degree of stability of the system. Therefore, this method does not apply to every changed condition appropriately and quickly in planning and operating. And Transient Energy Function (TEF) method whis can assess quickly and quantatively the degree of stability of the system and which judges the stability and the instability to analyse transient dynamic charater of the system by mutual changing kinetic energy and potential energy, is developed. TEF method analyses the first Swing transient stability of the system by using the thought that if after disturbance happening, the increase of all the rotator kinetic energy changes into the potential energy after diturbance clearing, the system is stable, otherwise the system is unstable. This paper represents the availabiIity of the TEF method by comparing with time simulation method on the two cases.

An approximate method for the dominant eigenvalue of one machine connected to the infinite bus has been suggested. This method is based on combining the traditional eigenvalue sensitivity analysis and the structure of the system state matrix. Numerical examples are presented. This method is considered to be quite useful in the stability analysis for various initial conditions and for adjustment of generator controller parameters.

In operation of power systems, operators must solve overloads in abnormal conditions instantaneously for restoring the power system to normal state. Earlier works proposed the power flow distribution and the load shedding decision, developed by the expert system, for settling overloads. This paper presents a method for solving special conditions in the earlier works which requires less amount of loads than that of the load shedding, in particular, of the load shedding considering the priority. This paper uses the effort system to improve the ability of a beginner as well as of trained operators for operating the power system during emergency operation. To prove efficiency of the proposed method, this paper has adopted the model system with 13 bus for obtaining simulation results and used the prolog language to implement the system.

The ADS system is applied and used in KEPCO. But it is short of graphic functions and on-line display of power system. In this study, we develpoed ADS MMI system which expands graphic functions of ADS, and enhanced efficiency of ADS system. And we showed system extension in the future and suggested a foundation of total distribution automation system.

In an automated power distribution systems, the fault situation is controlled from the control center through remotely-controlled sectionalizing switches. In this paper, the service restoration strategy which can yield the minimal switching actions and constraint checking is proposed and an expert system which generates the appropriate switching actions is described.

This paper proposes the methods to construct the tree-structured database and analyze the distribution system network. In order to cope with an extensive amount of data and the frequent breaker switching operations in distribution systems, the database for system configuration is constructed using binary trees. Once the tree-structured database has been built, the system tracing of distribution network can be rapidly performed. This remarkably enhances the efficiency of data search and easily adapts to system changes due to switching operations. The computation method of fast power flow using tree search strategies is presented. The methods in the paper may be available in the field of distribution system operation.

This paper presents results of typical sutdy on lightning overvoltages of underground distribution systems and discusses effective insulation coordination schemes. Typical lightning current and system parameters are used to analyze lightning overvoltage distribution on underground distribution systems by EMTP (Electro-Magnetic Transients Program).

In spite of recent stagnation, the rate of electric consumption has highly grown in accordance with a high cultural livings, Also, a usage of electric energy is growing with development of multi-information society. To make the matter worse, hard work of highest order is required because distribution system are widely spread and complex. And consummers demand for the electricity of good quality and it is obstracle to be not able to train related technitian for maintenance without an interruption of electric power. It is sure the counterplan of uninterruption of electric power is very necessary. In this paper, considering the working circumstances, we presented the necessity of work method, the effects and the future prospect for uninterruption power supply.

The distribution automation functions include feeder automation, data acquisition and processing, remote meter reading and load management. This report explains the function of feeder automation and the accomplishment method of feeder automation using the switches of distribution system with multi functions.

IKL Map has been drawn by the contours of the equivalent thunderstorm days for 23 years(1968-1990) in order to apply this map to the designs nod maintenances for lightning protection on the electric equipments.

This paper introduces an expert system for the automation of power system restoration. We focus on following two points : a) Data Structure which can describe power system and ensure the efficient search and b) Restoration Strategy which performs faster black-out restoration with fewer switching operations and overload alleviation using sensitivity. Some examples are used to illustrate the capabilities of the proposed expert system.

Snow accretion on transmission lines may cause critical accidents such as sleet jump, tower collpases, etc,. In order to assist power system operators to prevent thses accidents, aprototype expert system had been developed. The system has basic functions of forecasting snow accretion on transmission lines and making a list of all feasible and effective deicing countermeasures. The knowledge of the effect of weather conditions on the snow accretion process continue main rule base. Simulation results show that the proposed system is very promising.

Transmission loss retains a complete set of system state information. If we could exploit its behavior, we could find various useful applications. This study is a such exploitation to find same possibilities of the loss application. It seems to have very prospective possibilities. It has been tried to establish some network indices to indicate operation margin. And further some prospective applications are suggested.

In the conventional load flow technique, it is assumed that the generator at the slack bus is used to supply the transmission losses and the change of power due to the generator outage. The assumption is not true in physical sense. This paper presents a new load flow technique that considers the governor-frequency characteristics and load-frequency characteristics and the technique is consistent with the actual power system phenomenon. This paper proposes an efficient methodology using sensitivity with the new technique for contingency analysis, which is used to calculate the line flows. Computational results of this technique applied to IEEE 14-bus system are presented.

The power flow calculations are the most important and powerful tools in the various studies of power system engineering. Newton-Raphson method, among the various power flow calculation techniques, is normally used due to its rapidness of numerical convergency. In the conventional Newton-Raphson method, however, there are some unrealistic assumptions, in which all the system power losses are considered to be supplied by the slack bus generator. Introducing the system power loss formula and augmenting the conventional Newton-Raphson power flow method, we can relieve the unrealistic assumption and improve the performance of power flow calculation. In this study, A new approach for handling the losses and augmenting the conventional power flow problem is proposed. The proposed method estimates the increamental changes of active power on each generation bus with respect to the change of total system power losses and the estimated value are used to update the slack bus power. If some studies for more theoritical investigations and verifications are followed, the proposed approach will show some improvement of the conventional method and give lots of contribution to increase the performance of power flow techniques in power systems engineering.

Fuzzy L. P. with Multi-criteria function is adopted in this VAR allocation algorithm to accomplish the optimization of co-conflicting objectives, such as the amount of the VAR Installed and power system loss, while keeping the bus voltage profile within an admissible range. fuzzy L. P., a powerful tool dealing with the fuzziness of satisfaction levels of the constraints and the goal of objective functions, enables us to search for the solutions which may contribute in VAR planning. This advantage Is not provided by traditional standardized L. P. The effectiveness of the proposed algorithm has been verified by the test on the IEEE-30 bus system.

This paper introduces a new methodology to apply the interpolation technique wi th optimum pattern to voltage-reactive power control of power system. The conventional tool for the optimal operation of power system is Optimal Power Flow(OPF) by standard optimization techniques. The achievement of solution through OPF programs has a defect of computation time, so that it is impossible to apply the OPF programs to the real-time control area. The proposed method presents a solution in a short period of time and an output with a good accuracy. The optimum pattern is a set of input-output pairs, where an input is a load level and a type of outage and an output is the result of OPF program corresponding to the input. The output in the OPF represents control variables of voltage-reactive power control. The interpolation technique is used to obtain the solution for an arbitrary input. As a result, the new technique helps operators in the process of the real-time voltage-reactive power control in both normal and emergency operating states.

Recent years voltage collapse phenomenon have a great attention to power system engineers. As the system size increases the voltage problem shows a very complicated and the reactive power contol problem becomes more difficult. This paper gives an efficient methods for calculating voltage collapse proximity index based on the reactive power loss sensitivity and real power loss sensitivity. The system voltages are tightly associated with the system reactive power, so the proposed voltage collapse proximity index is very usefull for the system voltage control problems. Numerical examples showed a good and reliable results.

In this paper, a new learning fuzzy logic controller(LFLC) is presented. The proposed controller is composed of the main control part and the learning part. The main control part is a fuzzy logic controller(FLC) based on linguistic rules and fuzzy inference. For the learning part, artificial neural network(ANN) is added to FLC so that the controller may adapt to unknown plant and environment. According to the output values of the ANN part, which is learned using error back-propagation algorithm, scale factors of the FLC part are determined. These scale factors transfer the range of values of input variables into corresponding universe of discourse in the FLC part in order to achieve good performance. The effectiveness of the proposed control strategy has been demonstrated through simulations involving the control of an unknown robot manipulator with load disturbance.

A new algorithm is proposed to identify the structure and the parameters of the nonlinear discrete-time plant with only the unknown dynamics and the weak informations about its structure. The proposed algorithm is constructed with the compensation method of weghing values using its previous derivatives and with the efficient technique updating self-learning coefficients. The result in this application is thought to prove the effectiveness of the algorithm proposed in this paper and its superiority to the conventional ones.

This paper proposes a new algorithm to estimate time-varying parameters by combining KFSM(Kalman Filter with Shift Matrix) with neural network compansator. While the time varying parameters are estimated from KFSM, the error coverence of system, R(k) are compansated by neural network concurrently. The casestudy using computer simulation proves the usefullness and advantages of the proposed algorithm in this paper.

In this paper, back-propagation neural network is used for the identification and trajectory control of a DC motor. The neural network is trained to identify the unknown nonlinear dynamics of the motor and load and the trained neural network is used for speed control of the DC motor to have good performance. Simulation results show the good performance of the control system based on the neural network under arbitrarily chosen speed trajectories.

Conventionally in the industrial control, PlD controller has been used because of its robustness, and nonlinear characteristic of a system under control. Although the PlD controller produce suitable parameter of the each system and also variable of PlD controller should be changed according to environment, disturbance, load. In this paper, the convergence and learning accuracy of the back-propagation(BP) method in neural network are investigated by analyzing the reason for decelerating the convergence of BP method. and examining the rapid deceleration of the convergence when the learning is executed on the part of sigmoid activation function with the very small first derivative. The modified logistic activation function it proposed by defining the convergence factor based on the analysis and applied to the position and speed control of a DC-servo motor. This paper revealed for experimental, a neural network and a PD controller combined off-line system using developed the position and speed characteristics of a DC-servo motor.

We utilize as a processor TMS320C25 (Texas Instrument) in making a driver for a 4 pole PM synchronous servo motor. TMS320C25 has a 32bit ALU and a 16 bit hardware multiplier, and the maximum instruction execution rate is 10MIPS at 40MHz. We adopted a space vector modulation PWM method. An interesting point of this work is that PWM wave is generated by utilizing timer interrupts. Hence, in the rest of time the processor can take care of the other routine such as Park's coordinate transformation and the computation required in the feedback loops. Thus, it mates the hardware circuit very simple. Due to the decrease in the number of components, the motor drive system becomes more fault-tolerant and cost-optimized. Also, more flexibility is gained in changing the control parameters.

In this paper, the RedMAP, i.e. a redundant Mini-MAP system for high reliability is proposed. Redundancy is implemented for LLC, MAC, and Physical layer of ISL-Mini -MAP. The detection of error of the network, the broadcasting of the error event, and the network change sequence are three major functions for the dualized Mini-MAP system. The abnormal operation of the network is mainly detected indirectly with the function of the TBC( token bus controller). The time delay to be required for the change of the networks must be minimized. With the RedMAP, we can achieve successful transmission only with short additional recovery time.

This paper proposes a new algorithm that finds fuzzy control law of the system in which little knowledge has been known. In view or conventional fuzzy method, making control law needs the sense and the knowledge of the system which are provided by expert. But fuzzy control using proposed algorithm needs no expert for hating control law. After construction of the 1st order approximated ARMA model using input-output pairs, new defuzzification method is applied. The deduced rule is stored in fuzzy input space and updated by the proposed algorithm adaptively. To show the validity and effectiveness of proposed control method. simulation result is presented.

The expert system described in this article tunes a proportional-integral-derivative(PID) controller for a single-input single-ouput process. The expert system examines features of each transient response and the corresponding controller parameters. It determines a new set of controller gains to obtain a more desirable time reponse using fuzzy logic. This technique can be used to determine and implement a different set of PID gains for each operating regime and, once in steady state, the system can be used to find optimal parameters for load disturbance rejection.

Existing methods for controlling systems containing unknown dead-zones show the limitations on the performances. To solve these problems, this paper proposes a novel two-layer structured Fuzzy controller by analyzing the output response of the existing Fuzzy PD controller and the precompensator.

In designing the fuzzy rule-based system, it has effected by the four significant factors such as the choice of membership function, scaling factor, the numbers of fuzzy control rule, the method of defuzzification. In this paper we design the fuzzy rule based system and evaluate by three factors, as followes reaching time, overshoot, and amplitude. And then we wiII show that the significant factors are the choice of scaling factor and the numbers of fuzzy control rule, and the system performance can be improved by the proper selection of the scaling factors.

The control performances of a fuzzy controller depend on its control rules, I/O membership functions, and scaling factors. Scaling factors are very important to adjust control parameters to the process which is to be controlled. For tuning the sealing factors, trial and error method is used in conventional fuzzy controller, which is very difficult and time consuming. This paper proposes a tuning method of scaling factors based on the concept of tuning rules for the conventional Pl controller parameters. The proposed automatic tuning fuzzy controller was evaluated by computer simulations. Good results have been obtained for the small hydro power plant.

In this paper, an implementation approach of Fuzzy Position Controller for DC servo motor which requires the faster and more accurate dynamics is presented. The proposed Fuzzy position controller is consists of an adjustment routine obtaining optimal scale factors and a Fuzzy control routine inferencing optimal control signals, using 80286 microprocessor and DT 2801 board. Comparing to conventional PD controllers, the control performances such as reaching time, overshoot, disturbance adaptability and etc. are substantially improved.

A dynamic system containing uncertain elements is considered. The upper bound of the values of these uncertainties is estimated. Using the estimated value a bounded error state feedback control is proposed based on Corless and Leitmann's result [1].

This paper presents a method of hierachical optimal control for time invariant large scale systems via Single Term Walsh Series. It is well known that the optimal control of a large scale system with quadratic performance criteria often involves the determination of time varying feedback gain matrix by solving the matrix Riccati differential equation, which is usually quite difficult. Therefore, in order to solve the problem, this paper is introduced to Single Term Walsh Series. The advantages of proposed method are simple and attractive for the control of large scale system in computation.

In this paper, an LQG/LTR procedure for stable nonminimum phase systems is suggested using predictor scheme. In the method, the performance of the target feedback loop can be easily adjusted and the recovery error is less dependent on the location of NMP zeros than previous methods. The gain and phase margin and the robust ness for modeluncertainty of the suggested control system are obtained.

In this paper, we deal with design method of two-degree-of-freedom control system which desired property of robustness and tracking can be achieved simultaneously, Controller is designed by means of model matching method and H$\infty$ weighted sensitivity minimization design method. Satisfactory result of design example is obtained by simulation.

This paper considers the design of robust stabilizing controller of a linear time-invariant digital system subject to variations of parameter vector. For a given controller the radius of the largest stability hypersphere in this parameter space is calculated. This radius is a measure of the stability Margin of the closed-loop system. Based on this calculation a design procedure is proposed to robustify a given stabilizing controller. This algorithm iteratively enlarges the stability hypersphere in parameter space and can be used to design a controller to stabilize a plant subject to given ranges of parameter perturbations. These results are illustrated by an example.

The method of simulated annealing is a technique that has recently attracted significant attention as suitable for optimization problem of very large scale. If the temperature is too high, then some of the structure created by the heuristic will be destroyed and unnecessary extra work will be done. If it is too low then solution is lost, similar to the case of a quenching cooling schedule in the Simulated Annealing (SA) phase. Therefore, a crucial issue in this study is the determination of the starting temperature and cooling schedule for SA phase.

This Paper presents the design of an automated EEG analyzing system. The design considerations including processing speed, A/D conversion, filtering, and waveforms detection, are overviewed with the description of the associated EEG charateristics. The architecture of the currently implemented system consists of a -controller based front-end signal processing unit and a host computer system. The data acquisition procedures are described along with a couple of illustrations of the acquired EEG/EOG signal.

Presented in this paper is a generalized norm bound for the continuous and discrete algebraic Riccati equations. The generalized norm bound provides a lower bound of the Riccati solutions specified by any kind of submultiplicative matrix norms including the spectral, Frobenius and $\ell_1$ norms.

The adaptive active noise control algorithm for duct noise attenuation is considered. A new algorithm to estimate the secondary path transfer function using multiple models is presented. The computational burden of the proposed algorithm is much smaller than the existing methods, so it could be applied to the multi-channel cases. Computer simulations were done to show the effectiveness of the proposed algorithm in a duct case.

This paper presents the method to reduce the order of high-order linear time invarient system via Walsh function. It is based on the matrix pseudoinverse algorithm to determine the parameters of the reduced model which minimize the sum of the squares of the errors between the reponses of the high-order system and a reduced model to a given input. This proposed method can be conveniently implemented with a computer. They will be very useful in the study of control system via Walsh function.

This paper presents a technique, which identifies both process noise covariance and sensor noise covariance by using innovation correlation test. A correlation test, which checks whether the square root Kalman filter is workingly optimal or not, is given. The system is stochastic autoregressive moving-average model with auxiliary white noise Input. The linear quadratic Gaussian control is used for minimizing stochastic cost function. This paper indentifies Q, R, and estimates parametric matrics $A(q^{-1}),B(q^{-1}),C(q^{-1})$ by means of extended recursive least squares and model reference control. And The proposed technique has been validated in simulation results on the fourth order system.

Adaptive noise cancelling using system identification techniques for cancelling power line interference in the electrocardiogram (ECG) is presented. This method is sensitive and self-adjusting to both slow and abrupt changes in the AC interference amplitude and frequency. Computer simulation were done to compare this method with the Lekov's method.

Automated analysis of EEG invariably requires the inclusion of a signal conditioning filter. This paper investigates the EEG waveform distortions caused by the transient responses of the various types of signal conditioning filters. This study explicitly simulates the filter responses to the typical EEG waveform models, and compares the distortions.

A method is developed for the detection of parametric change for conditionally-linear stochastic systems. Such systems are quite prevalent in biology, economics and engineering, and may be synthesized as certain bilinear stochastic systems with output feedback (possibly nonlinear) through the controls. In other words, these systems are linear in the "unmeasured states" and nonlinear in the "measured states." The conditionally linear filter developed by Kolodziej forms a convenient part of the algorithm which estimates the time of change of parameter values.

The proposed algorithm uses a modified adaptive law which consists of switching -modification, normalized augmented error and low-pass filtered signal of output tracking error, furthermore, the proportional term that is a product of the output tracking error and the bounded signal having an information of output tracking error is added to the conventional control law for improvement of robustness and performance of an adaptive system. For the arbitrary nth order system, mathematical analysis and computer simulation are used to demonstrate improvement of output error characteristics, guaranteeing boundedness of all signals in the overall system.

This paper proposes indirect adaptive pole placement adaptive controller using a modified least squares method. If an adaptive controller has good performance, it is necessary that an estimator have fast convergence. This paper presents a modified least squares method which guarantees the stability of estimator and has fast convergence. In this algorithm, information on signal level is obtained from the determinent of covariance matrix and according to it, weighting factor is tuned.

In this paper, the active noise control(ANC) system using the inverse modeling techiniques is presented. The nonlinearity and time delay of the transfer function from the secondary speaker to the error microphone makes the ANC system have poor performance. To solve this problem, the inverse model technique and filtered-X LMS algorithm is using proposed. This proposed ANC system is implemented using DSP chip and operated in on-line. The experimental results show that this ANC system has better noise canceling performance than that used LMS only about 5-15[db]

This Paper presents an algorithm for improvement of convergence in Adaptive Filter. The proposed algorithm employes the time-varying Convergence Factor to Block LMS adaptation algorithm. Computer simulation for Adaptive Modeling of time-invarying and time-varying unknown systems has been performed. Performance comparisons between LMS, BLMS algorithms which have fixed Convergence factors obtained by trial and error and the proposed algorithm which has time-varying Convergence Factor show that the proposed algorithm is superior to LMS and BLMS algorithm with respect to speed and accuracy of adaptation.

This paper tries to evaluate the control performance of systeme in cases of parameter perturbations. Six cases of the root distribution and location changes of each characteristic roots by the parameter perturbation are considered as evaluation factors. A qualitative evaluation is performed through several simulations. These results will be used as a basic data for the complete analysis of the control performance.

In this paper, We propose a PID-type of self-tuning algorithm which is based on the parameter estimation and the minimization of the cost function. We use the CARIMA model for parameter estimation and determine the discrete PID controller parameters by minimizing the cost function which considers the quadratic deviations of the predicted output over the set-point as well as the control efforts. Also, The algorithm is extended by incorporating constraints of the control signal. Simulations are performed to illustrate the efficiency of the proposed method.

This paper deals with the tuning method of PID controls for process controls. It introduces the normalized process gain and the normalized process dead-time for processes based on Ziegler-Nichols tuning methods. In the case of PID auto-tuning, the first, this method applies Ziegler-Nichols tuning method and introduces the set-point weighting for reducing overshoot in the large normalized process gain or small normalized process dead-time, the second, this method is modified and includes the set-point weighting in the small normalized process gain or large normalized process dead-time. In the case of PI auto-tuning, this method is modified for reducing overshoot. This paper obtains empirical data with Ziegler-Nichols methods for refined Ziegler-Nichols tuning methods.

In this paper, To estimate monthly water distribution volume required optimization control of operating scheme & water distribution management for water transmission system in water supply, both Thomas-Fiering technique and Fourier series are compared and analyzed, respectively. Since water distribution volume is periodically repeated and has a linear fluctuation trend, parameters in each element are estimated through dividing into linear fluctuation trend component and periodical component. Finally, results of time-series analysis are proved to be more reasonable than that of Thomas-Fiering techniques by comparing simulation with observation data.

In this paper a robust hybrid control algorithm for n-link nonredundant robot manipulators is proposed. This scheme includes an estimation law for the upper bound on the uncertainty such that robust control input is updated as a function of the estimated upper bound. The uniform ultimate boundedness of the tracking error is generated by the Lyapunov based theory.

In this note, we propose a method for designing a robot controller which can suppress the effects of both the model uncertainty and noisy velocity measurements. The controller is an output feedback compensator of which the constant gains are given in terms of a Riccati equation and a Lyapunov equation. The controller guarantees not only uniform boundedness but uniform ultimate boundedness. The stability result is local but the region can be arbitrarily enlarged at the expense of large control gain. The control law needs neither the exact knowledge of the physical robot parameters nor clean velocity measurements.

Recently, robot plays an important role in factory automation and its use is rapidly increasing. In paticular, modern industry tends to need the robot to adjust itself swiftly to new workcell. But, robot is to be taught of operator through teaching box in order that it can carry out new tasks. In the process, interruption of the production line is frequent and so much time and expense are required. Computer Graphic Simulator(CGS) provides a useful tool to overcome such a problem. In this paper, CGS consists of teaching mode and execution mode. The 5-axis Rhino robot is modelled in IBM-PC/386 compatible 32-bit personal computer with 80387 co-processor and the simulator performs defined operations on the computer screen.

본 연구에서는 n-차 비선형 로보트 매니퓰레이터에서 링크의 동특성에 액류웨이터의 동특성을 포함시킨 복합모델(Combined Model)의 알고리즘을 제안하였다. 여기서 이 복합 모델의 알고리즘은 고속운전 및 빈번한 부하의 변동시에 중요한 역할을 하는 액류웨이터 모터의 동특성 방정식과 매니퓰레이터 링크의 운동을 묘사하는 동특성 방정식을 포함시킨 알고리즘을 의미한다. 또한 제안된 모델은 선영궤환 및 감결합 변환(Feedback Linearizing and Decoupling Transformation : FLDT)법을 이용하여 선형시스템으로 변환시켰다. 따라서 본 연구에서는 고속운전 및 빈번한 부하의 변동에도 액류웨이터 모터의 전압과 전류를 조절함으로써 효과적으로 제어할 수 있게 하는 매니퓰레이터 모델 알고리즘을 개발하는데 그 목적이 있다.

This paper describes the development of an Integrated Digital Protection And Control System (IDPACS) for 345 KV substation. The developed IDPACS system containes digital protection units for transmission lines, delivery lines, transformers, bus, shunt reactors and Shunt capacitors. We deloped a Local Area Network (LAN) Which connects the digital protection units to the station computer for monitoring and control purposes.

In this paper, the development of a fiber optic network for an integrated digital protection relay system is described. The structure of the developed network is determined to loosen the optic requirements and to have good extensibility while providing sufficient functions for protection and control for substations. The network has a hierarchical structure with two levels. The upper level handles data for monitoring and control of the system with star topology. The lower level manages the real time data for bus protection with one-to-one connections. Communication flows of each level are described. The HDLC is employed as the network protocol.

This paper presents an approach to the design of real time task scheduling for a line controller of continuous manufacturing process automation. The line controller has multiprocessor-based architecture with shared memory and is operated by firmware. This firmware contains menu-driven software supporting real-time database management and fuction-block control language. The multitasking line control processor performs the following three functions: 1) interprets the function block control language by virtue of shared memory in the database; 2) invokes an interupt service routine as required by external hardware; 3) detects errors and notifies the user. We propose real time task scheduling method.

This note proposes a novel robust adaptive control algorithm for systems with unknown disturbances by introducing an additional term in the control input. This additional term is easily implementable by estimating the upper bound of the unknown disturbances. By this term, the output error can be made to be uniformly ultimately bounded in a desired region via Lyapunov second stability theorem when the relative degree of system is one.

This paper proposes a stable locomotion control rule for non-holonomic mobile robot. Stability of the rule is proved through the use of a Liapunov function. We have two controller for locomotion control. One is velocity controller, the other is position controller. The proposed controller is position controller whose input to robot are a reference posture and reference velocities. The major objective of this paper is to propose a control rule to find a reasonable velocity command under a assumption which is velocity controller is ideal controller.

This paper proposes a novel direct adaptive pole placement control algorithm which can be applied to continuous time nonminimum phase systems. The algorithm is based on Lyapunov's direct method. By introducing an auxiliary signal, a minimal error model is constructed in state space. Using the error model an estimation law is obtained via Lyapunov's second stability theorem. The global stability of the overall system is established.

In this paper, dynamic modeling and adaptive control problems for a space robot system are discussed. The space robot consist of a robot manipulator mounted on a free-floating base where no attitude control is applied. Using an extended robot model, the entire space robot can be viewed as an under-actuated robot system. Based on nonlinear control theory, the extended space robot model can then be decomposed into two subsystems: one is input-output exactly linearizable, and the other is unlinearizable and represents an internal dynamics. With this decomposition, a normal form-augmentation approach and an augmented state-feedback control are proposed to facilitate the design of adaptive control for the space robot system against parameter uncertainty, unknown dynamics and unmodeled payload in space applications. We demonstrate that under certain conditions, the entire space robot can be represented as a full-actuated robot system to avoid the inclusion of internal dynamics. Based on the dynamic model, we propose an adaptive control scheme using Cartesian space representation and demonstrate its validity and design procedure by a simulation study.

Algorithmic or kinematic singularities are inevitably a introduced if optimality criteria or augmented kinematic equations are used to resolve the redundancy of almost any manipulator with rotary joints. In this paper, a sufficient condition for a singularity-free optimal solution of the kinematic control of a redundant manipulator is derived and, specifically, algorithmic singularities are analyzed for optimality-based methods. A singularity-free space (SFS) to characterize the performance of a secondary task for a redundant manipulator using the sufficient condition for a redundant manipulator is defined. The SFS is a set of regions classified by the loci of configurations satisfying the inflection condition for manipulability measure in the Configuration space. Using SFS, the topological property of the Configuration space and the invertible workspace without singularities are analyzed.

This paper proposes an adaptive control system using a 80286 microprocessor-based system and DC servo motors for the control of flexible joint manipulator. In this paper, we construct the controller based on a singular perturbation strategy damping out the elastic oscillations at the joints. we added to the controller the compensator for damping the joint and the term for decreasing the position error between the actuator and the link in order to improve the asymptotical convergence of the position of the link. It is shown that the implementation of this control algorithm can be practical.

This note is concerned with the point to point control of manipulators having elastic joints. We present a PD control algorithm which is adaptive with respect to the gravity and elastic parameters of robot manipulators. While the conventional control law is used, a new adaptive law is used to improve the performance. The proposed controller is shown to be stable. It is Shown that steady-state position error converges to zero through some simulations concerning the manipulator with three revolute elastic joints.

Many dynamic control have been proposed; however, most of them are limited within stage of simulation study. The main reason is that the computations required for inverse dynamics are far beyond the ability of the present commercially available microprocessors. In this paper, In order to achieve real-time processing in robot dynamic control, a parallel processing computer for robot dynamic control is implemented using two transputer. Two transputer compute two degree of freedom robot. The transputer is a special purpose MPU for parallel processing. Transputers are used in networks to build a high performance concurrent system. A network of transputers and peripheral controllers is constructed using point-to-point communication. To gain most benifit from the transputer architecture, the whole system is programmed in OCCAM which is a high level language for concurrent applications. This control algorithm is applied to the RHINO SCARA type manipulator. We could taked about 438.6 microseconds to compute robot dynamic with two-processors.

For the control of robotic manipulators with unknown or uncertain dynamics, leaming control schemes are very effective control schemes for repeated trajectory following tasks. In this class of controllers, control techniques using neural networks have been gaining much attention in recent years.. In this note, we discuss the leaming control techniques using neural networks and propose a new model-based control scheme using multilayered neural networks. Three-layerd neural network is used as a feedback controller to compensate the mismatched terms between model plant and real plant. Computer simulations are performed to show the applicability and the limitation of the proposed controller.

In this thesis, the algorithm of Fuzzy Logic Control(FLC) is applied to the Nonlinear system to implement a system response. Proportional-Integral-Derivative(PID) controller is also used to control the various systems. Look-up table is applied to decide the control input, and the other look-up table is added for saving memories and inference time. Generally, FLC input variables are error(E) and error derivative($\Delta$E). In this algorithm. another Input variable error's second derivative($\Delta^2$E) is added for Robust Fine control.

An adaptive hybrid position/force controller for constrained manipulator with uncertain dynamic model parameters and environment stiffness is presented. In this paper, the compliance frame model is constructed by independent positions and forces to be controlled. The adaptive controller based on this compliance frame dynamic model is designed. Lyapunov theory is used for controller design and Stability analysis.

This paper proposes a simple and high performance hybrid position/force control of robots based on disturbance compensation by using the disturbance observer in task coordinate space. The disturbance observer linealizes system of robot manipulators in task coordinate space and realizes acceleration control. To realize the strict acceleration control, the disturbance observer whose input is a position signal by simple computation, works as if it were a disturbance detector. The inverse kinematics can be simplified, because the disturbance observer in task coordinate space compensates not only the disturbance but also the error due to the simplification of the inverse kinematics. The new strategy is applied to a three-degrees-of freedom direct drive robot. The robust and simple hybrid position/force control is realized experimentally.

두팔 robot의 dynamic model 에 위치 및 힘의 제어를 적용했을때의 부하분배의 관한 연구가 논문의 주요내용이다. 두팔 robot의 완전한 control을 위해서는 위치 뿐만아니라 힘이 제어를 하여야하며 그럴때의 전제 동력학 식의 타당성 여부를 알기위해서는 안정도 연구를 하여야 한다. 그리고 위치 및 힘의 제어를 했을때의 Load의 distribution 문제에 대하여 언급했다. 이에 관한 연구도 연재 많이 진행되고 있으며 여기서 소개하는 방법은 Load 및 end-effector의 힘의 분포에 주안점을 두어서 해석을 하였다. 이 경우에 Load의 분포에 관한 최적해를 얻기위해서는 두팔에서 소모되는 에너지의 최소치를 찾는 방법과 또 하나는 Load에서 형성되는 힘의 기하학적 구조에 대한 2가지 해석을 하였다. Load의 힘의 분포를 해석하기 위해서는 Load의 동작에 무관하게 적용되는 internal force를 가정하였다.

Ultrasonic and vision data need to be fused for efficient object recognition, especially in mobile robot navigation. In the proposed approach, the whole ultrasonic echo signal is utilized and data fusion is performed based on each sensor's characteristic. It is shown to be effective through the experiment results.

In this paper, we are designed a hierachical system controller and builed a robot system for high precision assembly consisting in multi-arms and multi-sensor. For the control of a multi-arms robot system, the robot system are consisted of cell controller, station controller and device. The Operating System of a cell controller is VxWorks for real-time multi-processing. Using by C-language, we are proposed a multi-arms robot control language based a RCCL, and this control language is partially implemented and tested in multi-robot control system.

In this paper, we proposed an algorithm for the musical note recognition. Firstly, a given bit-mapped music score image is converted to a set of individual note pattern images via vertical projection. Then, the pitch of a note is determinal by comparison in the note-head position with the reference five-lines. Also, the length of a note is found via leader clustering with a set of normalized note patterns. Finally, a datafile to play the music is obtained using the pitch and length of musical notes. Experimental results with a simple musical score image show that the proposed scheme is performed well.

In this paper, we propose a new multilevel nonlinear filter for simultaneous edge detection and noise suppression, which we call a order statistic-median hybrid(OMH) filler. The median-related filters cause an edge shift in the presence of an impulse near the edge. The proposed filter reduces such edge shifting while suppressing impulsive as well as nonimpulsive noise. We show that at the noisy edge point the OMH filter is substantially superior to the median filter, the $\alpha$-TM filter and the STM filter[I] in two respects: (a) the output bias error and (b) the output mean square error. Test results confirm that the OMH filter is robust in preserving sharp edges, inhibiting edge shifting, and suppressing a wide variety of noise. The structure for the OMH filter integrated circuit is also described.

We design a multi-element photo-detector to measure the size of particles using the diffracted light energy distribution. The light energy that is scattered by particles in the collimated laser beam is collected by the Fourier transform lens and directed to the multi-semicircular concentric annular detecters. The scattered profile measured by the photodetector is sampled by a 32 channel analog-to-digital converter. A nonnegative least squares analysis translates the light energy distribution into the corresponding unique particle size distribution.

In this paper, for tracking and filtering seaway information which act as a control disturbance the adaptive notch filter which removes disturbance with fast convergence and stability without changing the value of gain parameter $\mu$ when statistical property of input signal varies rapidly is designed by improving conventional RMXMS(Recursive Maximum Mean Square) algorithm. Besides, in consideration of measurement noise of sensors in underwater vehicle, the system which removes the noise and the disturbance is suggested.

In this paper, we contructed extended Kalman filter for the image navigation systems. The conventional extended Kalman filter methode are simulated for nonlinear measurement systems. In addition, we designed a maneuvering target tracking filter using Singer's model technique and input estimation technique by Chan. Simulation results show that Chan's input estimation technique has performed better than Singer's technique.

Before performing an adaptive stereo matching using satellite images, classification is needed as a preprocessing step. This paper describes that classification of three land cover types : river, mountain, and agricultural fields. We proposed that classification algorithm using pyramid structure and texture features. Results of applying the proposed algorithm to satellite image improved classification accuracy.

A Data-based Covariance Matrix(DCM) is introduced in the Eigenvector(EV) method, among subspace methods of estimating multiple sinusoidal frequencies from finite white noisy measurements. It is shown that the EV with the DCM can obtain the true. frequencies from finite noiseless data Some asymptotic results and further improvement on the DCM are also presented mathematically. Monte-carlo simulations are statistically conducted from the view-points of means and standard deviations in the EV's of DCM and Conventional Covariance Matrix(CCM). Simulations show a great promise for using the DCM, particularly for the cases of short data records, closely spaced frequencies and high signal-to-noise ratios.

This paper presents the new method to recognize the 2D patterns dynamically by rotating the input patterns according to the difference vector. Generally neural network with many patterns leads to various recognition ratio. The dynamic management of input patterns means that we can move pixels to desired locations controlled by the difference vector. We divide dual neural network model into two parts at learning phase, respectively. And then we combine them to construct the total network. Our model has some good results such that it has less number of patterns and reduced learning time. At present, we only discuss the four way movement of input patterns. The research for the complex movement will be fulfilled later.

In this paper, an extraction of depth intonation in pattern using neural network is presented. All the 3D images represent the depth information in grey pixels. This pixels which have analog values translated digital values. Because of the noise and distortion in pattern, we use the normalization in learning and recalling the patterns. Our method has eight direction vectors and slopes for pattern. Also, we use potential to obtain the mean slope and direction vectors of given 3D patches. The higher level of deduction finding the global depth information is also carried out by using neural network.

This paper shows 3D vision system based on halftone image analysis. Any halftone image has its own surface vector normal to surface patch. To classily the given 3D images, all the patch on 3D object are transformed to black/white halftone. First we extract the general learning patterns which represents required slopes and their attributes. And next we propose 3D segmentation by searching intensity, slope and density. Artificial neural network is found to be very suitable in this approach, because it has powerful learning quality and noise tolerant. In this study, 3D shape reconstruct using pyramidian model. Our results are evaluated to enhance the quality.

In this paper, we classified electromyographic(EMG) signal for prothesis control using neural network. For this study fast Fourier transform(FFT) with ensemble averaged spectrum is applied to two-channeI EMG signal for biceps and triceps. We used the three layer network. And a cumulative back-propagation algorithm is used for classification of six arm functions, flexion and extension of elbow and pronation and supination of the forearm and abduction and adduction of wrist.

This paper compares performance of neural and fuzzy truck backer-upper control systems. Conventional controllers require a mathematical model of how outputs depend on inputs. Neural and fuzzy control systems offer a key advantage over conventional control systems. They are model-free controllers. Neural networks learn a control process by examples (training samples). Fuzzy systems directly encode designer's experience as IF-THEN rules. For robustness test, we gradually removed training samples for the neural controller, and fuzzy rules for the fuzzy system. The errors increased laster in the neural controller than in the fuzzy system.

This paper presents a neural network method for tuning PlD controller of a time-varying process. Three gains of PlD controller are tuned for a certain desirable response pattern by back-propagation neural network. The neural network is trained using changes of output features vs. changes of PlD gains. But sometimes it needs longer training time and larger structure to train the correlation between the process and controller on entire region of the process. The difficulty in system identification is that the inverse function of the system can not be clearly stated. To cope with the problem, we do not train the neural network to respond correctly for the entire regions but train for only local region where the system is heading toward by training the neural network and tuning of the PlD controller. It may be trained for fine-tuning itself. Simulation results show that the adaptive PID controller using neural network trained in the local area performs remarkably for time-varying second order process.

Although the back-propagation(BP) algorithm is widely used for its simple structure and easy learning method, it has a drawback of slow convergence rate. In this paper, we propose an algorithm to improve this problem by manipulating the slope parameter of the activation function. The steepest descent method is used in learning the slope parameter, as in the case of weight. The simulation shows that the learning rates of the proposed algorithm is faster than the conventional BP algorithm.

To resolve the knowledge acquisition bottleneck in the expert systems, the connectionist expert systems have been proposed, which facilitate learning capability of neural networks. This paper is to modify Gallant's connectionist expert network so that it can be applied to more general problems : 1) The hidden nodes are added between the input nodes and an output node, so that the back propagation learning algorithm is used instead of perception based Pocket algorithm. 2) Inference engine is thus modified by modeling that a node may have uncertainties due to unknown inputs.

This paper described the development example of optical CT(Current Transformer) and optical PT (Potential Transformer) which is measurement system of current and voltage of electric power line using Electro-Optic effect and Magneto-Optic effect. We designed the optical CT of 22KV class and 154KV class and an optical PT of 22KV class. And now the filed testing is in progress in a KEPCO's substation.

Hourly load forecasting has become indispensable for practical simulation of electric power system as the system become larger and more complicated. To forecast the future hourly load the cyclic behavior of electric load which follows seasonal weather, day or week and office hours is to be analyzed so that the trend of the recent behavioral change can be extrapolated for the short term. For the long term, on the other hand, the changes in the infra-structure of each electricity consumer groups should be assessed. In this paper the concept and process of hourly load forecasting for hourly load is introduced.

In this paper, we described a evaluating technique of deposition status of LB films by quartz crystal. As a deposited material for this experiments, arachidic acid and stearic acid were used. We observed the deposition status of LB films on the surface of quartz crystal. The experimental results showed that Y type H films are deposited according to frequency shift, deposition ratio, and SEM photographs etc.

This paper deals with the experiment of CW YAG laser for processing. The YAG laser is spotted very small size beam compared with $CO_2$ laser having short wave-length. That is used broadly in material processing because of easy reaction to the materials, and the maintenance is very simple. The power delivery and focussing is done conventional optical components, and splitting beam is used many point stop-welding. In these studies, especial interest is nesessary to perpare for future technology. Our study aims to develop the YAG laser system and to accumulate design and construction technology. In basic experiments, we obtain the maximum output power of 50 W with the single elliptical cylinder.

This paper describes the optimization of design variables of SLIM using finite element method and SUMT(Sequential Unconstrained Minimization Technique). Thrust is taken as an objective function in order to maximize thrust under constant current drive, and seven independent design variables and nine constraints are chosen. As a result, $\tau$/g(pole pitch/airgap) and $\tau$/$d_{AL}$(pole pitch/aluminum depth) of good criteria in SLIM design are determined.

When it comes to the 'Leak Detector System' generally, our country has a large income 'Rotary Type Leak Detector' of foreign goods. The completed development of the 'Line Type Leak-Detector' system was produced to check Whether the containers for small and large on the filling line are auto defective. This system is applied to the filling package Processing during the production and contributed to inproving competiveness of domestic containers manufactures of all society of Industry. Also, high precision and realiablity, very compact size, low cost and Easy set-up etc. by checking the experimental data directly plan, Design and making for '1 Compression Head Control Leak Detector System'. This flexcible system can be equipped with multiple Compression heads depending on the requested prodution test rate and test precision.

Modern control theory has been developed day by day, but it hardly has been applicated practically to the various fields of industry, specially power electronic field, because there is no cost effective hardware that accommodates modern control theory. Thus, in our study, DSP based ultra high speed general controller which may be exclusively used for power electronic applications is developed.