제어로봇시스템학회:학술대회논문집
Institute of Control, Robotics and Systems (ICROS)
 기타
Domain
 Machinery ＞ Robot/Automated Machinery
1988.10

The realization of high quality robot control needs the improvement of computing speed of controller. In this paper, parallel processing method is considered for this purpose. A S/W algorithm for task scheduling is developed first, and then, an appropriate H/W structure is proposed. This scheme is applied to calculate inverse kinematics of PUMA robot. The simulation results show that the computing time when using three 8086/87's is reduced to 4.23 msec compared to 10 msec in case using one 8086/87.

Alightweight robot arm carrying a payload is modelled as a cantilever beam with a tip mass subjected to a high speed rotation. Equations of Motion, for modal control, are represented as discrete state variable form. Digital optimal control law with observer is developed to suppress the arm vibration and control the position of the joint angle. The effects of the number of controlled modes, weighting factors of the performance index, reference rotation time, and sampling time on the control performance are analyzed by computer simulation and experiments.

A controller described in this paper is designed for implicit generalised minimum varience controller with variable time delays in which the weighting polynominals are calculated to reduce the output and control signals variances. This paper is based on the fact that the poleassigment equation may have multiple solutions if the weighting polynominals are not of minimal order. It is shown that the larger order of the weighting polynominals increment the better is the stochastic behavior of the closedloop system with variable time delays without changs in the deterministic behavior of the system. Based on this theory, the controller is applied to position control of a threelink manipulater with parameter uncertainty.

This paper presents a new algorithm for solving the inverse kinematics in realtime applications. The endtip movement of each link can be resolved into the basic resolution unit, .DELTA.l, which depends on link length, reduction ratio and resolution of the incremental encoder attached to the joint. When x and yaxis projection of the endtip movement are expressed in .DELTA.l unit, projectional increments .DELTA.x and .DELTA.y become 1, 0 or I by truncation. By using the incremental computation with these ternary value and some simple logic rules, a coordinate transformation can be realized. Through this approach, it should be noted that the floatingpoint arithmetic and the manipulation of trigonometric functions are completely eliminated. This paper demonstrates the proposed method in a parallelogram linkage type, twolink arm.

The application of robot to industry is increasing and as a result the study on robot is widely being carried out. In this study, to improve the accuracy of robot motion the method which calibrates initially assumed link parameters is considered. This method calibrates 4N link parameters for N D.O.F. robot with rigid links.

A study has been carried out on the implementation of IMCA (Linear Moving Voice Coil Actuator) to a flexible robot arm modelled as cantilever beam. Control performances are evaluated by computer simulation and theoretical analysis is validated by experiments. From this study, it is proved that the LMVCA can be applied easily to the control system and suppress vibration effectively.

Generally, for nonlinear control of robotic system, the method of computed torque or inverse dynamics is frequently used. In this case, exact knowledge of the system parameters is required, however. This paper addresses the problem of nonlinear control when the parameter of system is varied. The approach is based upon decoupled model following. As an example, control of a three degree of freedom manipulator arm under mass variation is simulated.

It is wellknown that the redundancy can be exploited to avoid the singular regions of the redundant manipulator by increasing the manipulability. The method, however, requires exprecessive energy and gives rather large tracking errors since the manipulability is increased repidly so that the manipulator avoid the singular region quickly. In this paper, a new method is proposed in which the increasing speed of the manipulability is confined to a certain bound. Therefore, in the proposed method, the movement energy and the tracking errors are reduced. The computer simulation studies are performed to show the validity of the method.

This paper represents the study of an effective selftuning PID control for a robot manipulator to track a reference trajectory in spite of the presence of nonlinearities and parameters uncertainties in robot dynamic models. In this control scheme, an error model of the manipulator is established, for the first time, by difference between joint reference trajectory and tracked trajectory. It's model Parameters are estimated by the recursive leastsquare identification algorithm, and classical controller parameters are determined by pole placement method. A computer simulation study was conducted to demonstrate performance of the proposed selftuning PID control in jointbased coordinates for a robot with payload.

This paper presents a centralized adaptive control scheme based on perturbation equations in the vicinity of a desired trajectory,which are used to design a feedback control about the desired trajectory. This adaptive control scheme reduces the manipulator control problem from a nonlinear control to controlling a linear control system about a desried trajectory. Computer simulation studies of a twojoint manipulator are performed on a IBMPC to illustrate the performance of this adaptive control scheme.

This paper presents an approach to the position control of a robot manipulator by using a selftuning poleplacement controller with an inverse model. The linearized independent difference equations of manipulator motion are obtained, and the parameters of these models are estimated on line. The controller is composed of two parts, the primary controller obtains desired torques by using an inverse model and the secondary controller computes variational torques on the basis of induced perturbation equations by minimizing a quadratic criterion with a closedloop poleplacement. Simulation is performed to demonstrate the effectiveness of this approach.

In this thesis, after the inverted pendulum being made, its state equation was established with parameters obtained by the state space approach and its position stabilization control was performed with VSS. The switching function of VSS was obtained from pole placement method and as the result of VSS control was compared with that of optimal regulator control, its robustness was clarified in sliding mode. The position and stabilization control of the inverted pendulum Wstem were acomplished with VSS.

Recently many adaptive control schemes for the industrial robot manipulator have been developed. Especially, learning control utilizing the repetitive motion of robot and based on iterative signal synthesis attracts much interests. However, since most of these approaches excludes the boundness of the input torque supplied to the manipulator, its effectiveness may be limited and also the full dynamic capacity of the robot manipulator can not be utilized. To overcome the abovementioned difficulties and meet the desired performance, we propose an approach which yields the effective learning control schemes in this paper. In this study, some stability conditions derived from applying the Lyapunov theory to the discrete linear timevarying dynamic system are established and also an optimization scheme considering the bounded input torque is introduced. These results are simulated on a digital computer using a threejoint revolute manipulator to show their effectiveness.

In the control of the robotic manipulators, the variable structure control method for the get Point Regualation has a advantage of the insensitivity about parameter variations and disturbances. When the robotic manipulators are controlled by a pointtopoint scheme, no path constraint is considered. Thus, the variable structure control method will be effectively applied only if the trajectory of the robot hand is estimated precisely. In this paper, the joint trajectories in the joint space and the hand trajectory in the cartesian space are calculated by the variable structure control method, and an algorithm is suggested to elaborate the deviation error of the robot hand from a straight line path. The result of this study will become a base of the effective path planning about robotic manipulators with the variable structure control concept.

We present an algorithm that uses trajectory following errors to improve a feedforward command to a robot in the iterative manner. It has been shown that when the manipulator handles an unknown object, the Ptype learning algorithm can make the trajectory converge to a desired path and also that the proposed learning control algorithm performs better than the other type learning control algorithm. A numerical simulation of a three degree of freedom manipulator such as PUMA560 ROBOT has been performed to illustrate the effectiveness of the proposed learning algorithm.

Due to the fact that the set point regulation scheme by the variable structure control method concerns only the initial and final locations of a manipulator, many constraints may exist in the application of path tracking with obstracle avoidance. The variable structure parameter should be selected in the trajectory planning step by satisfying the constraints of the travel time and the path deviations This paper presents the selection algorithm of the variable structure parameters with the constraints of the system dynamics and the travel time and the path deviation. This study makes unify the trajectory planning and tracking control using the variable structure control method.

As a means of human interface, this study designs DevelopedERES/WCS with voice recognition capability and fuzzy set theory. In the advanced teleoperator system, when an error occurs on the automatic mode, the error is recovered after the automatic mode is changed into the manual mode intervened by a human. The purpose of this study is to reduce human work load and to shorten error recovery time during error recovery.

This paper are the manufacture of controller of direct drive arm robot using 32 bit CPU(MC 69020). The work would draw on KIT of Robotics Laboratory whose extensive experience in 16 bit CPU Controller(MC 68008) in addition to the WHILE languages. We found that this controller is good for the direct drive arm robot controller for the use of selftuning algorithms and real time control.

The analytical studies and applicability of the implementation of the controller for the simple flexible arm studies and discussed in this thesis with the goal of developing flexible manipulator arm control. Minimumtime position control and quadracticoptimal control are investigated in this thesis. Its validity has been established using numerical simulations.

The path planning is done at the joint level. Cubic spline functions are used for constructing joint trajectories for industrial robots. For Njoint robot, Cartesian knots are transformed into N sets of joint displacements, with one set for each joint. For industrial application the speed of operation affects the productivity. An algorithm is developed to schedule the time intervals between each pair of adjacent knots such that the total traveling time is minimized subject to the physical constraints on joint velocties acceleration and jerks.

A simple look and move type navigation algorithm with obstacle avoidance was implemented on a mobile robot equipped with two cameras. The robot gets a difference of two images from each camera and detects the position represented on the horizontal line and width of each obstacle in a certain distance. The obtained informations of obstacles are used to select a next intermediate point and to plan a path to it. Robot motion is simplified to straight line motion and rotation.

A walking robot must have the ability to change the body direction in order to avoid the obstacles. In this paper, we develop a gait control algorithm that can maintain the stable movement of the robot for three different modes of changing directions. The algorithm makes it possible for the robot to have the larger gait stability margin than the threshold value by the method of changing the body speed.

다각 보행로보트를 효율적으로 운용하기 위한 총괄적인 관리제어 시스템이 설계되었다. 운영체제인 XINU를 시스템내의 프로세서간의 통신 및 작업 스케쥴링을 효휼적으로 하기위해 채용하였다. 보행로보트의 제어프로그램이 개발되었으며, 1각의 기구 모형에 대한 실험이 수행되었다.

The problem of controlling static gaits of a Quadruped Walking Robot is investigated. A theoretical approach to the gait study is proposed in which the static stability margins for periodic gaits are expressed in terms of the kinematic gait formula. The effects of the stride length to the static stability are analyzed, and the relations between the static stability and the initial body configurations are examined. Extensive computer simulations are performed to verify the analysis results.

In this paper, we developed an environment for the analysis of kinematic characteristics of a quadruped robot using animation with simple task planning in SUN3 workstation. The projection method and related graphic utilities are developed, and we considered user interface with mouse. The legs with pantograph mechanism are mathmatically modelled and they will be used for the simulation of algorithms to be develped in the graphic monitor.

This study presents a method to determine the ground reation forces of a quadruped walking machine on its foot ends caused by the body weight and the inertia forces from the commanded acceleration. The method shows the same result as the PseudoInverse Method when the 4 feet stand on a plane. However method can be applied even when the 4 feet stand on a nonplanar surface for which, no feasible solution can be obtained by the PesudoInverse Method.

In this paper we describe a motion control algorithm for a 4legged robot over slopped terrain and steps. The new concept of the mechanically constrained angle has been introduced and the algorithm has been developed based on the relationship between the gait stability margin and the slope angle. The result then has been extended to the case where the robot walks over steps.

A sampled data controller for the quadruped walking robot is presented. To provide systematic design procedure, the relation between PI gain of velocity controller and sampling rate is analyzed with the ISE performance index and the time responses. The position controller for oneleg, 3axis, was developed under consideration of compactness and expendability. And several experiments were performed.

Freepathtype guidance system does not need a hardwired path in the environment so that it gives a mobile robot a flexible path. ln this study to achieve the freepathtype guidance system for a mobile robot which is steered by the differential steering of both drive forewheels, position recognition systems are constructed using odometer system as an internal position sensor. Two odometer systems, a auxiliary wheel odometer and a 2encoder odometer system are constructed and path following algorithms using these odometer systems are designed and experimented. PID control type is adopted in the path following algorithms.

In this study, the stabilization of moving sight using a gyro is investigated. At first, Linear Compensator was design by linearizing gyro, torque motor and several parameters from a given required frequency response curve. By using this, System Control Performance was analyzed by back EMF, torque saturation and Coulomb friction effects. Also stabilization Performance by disturbances and Paramter variations were simulated.

In this paper, a hyperstable cascade control scheme for servo drive systems is proposed to reduce the chattering, which is an undesirable property in discontinuously controlled systems. First, a discontinuously controlled hyperstable MRAC scheme is designed with respect to PI speed control system using only output variables, and then a linear position controller is cascaded. It is shown that the above system is asymptotically stable and the chattering is greatly reduced at a constant speed provided that the disturbance torque satisfies a certain condition.

In this paper, the adaptive controller for the turret gun is discussed which uses model reference adaptive technique based on the Lyapunov direct method. Turret gun can be decomposed into two timeinvariant SISO control systems. One is for the elevation angle control and the other is for the azimuth angle control under the assumption of independence each other. Thus we only consider here about the control loop for the elevation angle.

This paper utilizes an optimal control law for the accurate tracking servo system design. The devivation of a simple control law implementing microprocessor is made to minimize position and speed error of the controller. The 16 bit microprocessor receives command angular position and calculate the control algorithm for accurate tracking and provides control system gain scheduling to achieve very short settling time. Simulation results and some experimental results of the position controlled tracking using 4.5Kw DC servo motor are shown.

This paper deals with the robustness property of Kalman filters for linear systems with delay in output. The operatortype Riccati equation is transformed to algebraic equations, and the circle condition is derived. Based on the circle condition, it is shown that the same nondivergence margin, (1/2, .inf.) gain margin and +60.deg. phase margin, is guaranteed as for ordinary systems.

In this paper, a new fast algorithm of FIR least squares filter with linear phase is presented. The general unknown statistics case is considered, whereby only sample records of the data are available. Taking advantage of the neartoToeplitz+Hankel structure of the resulting normal equation, a fast algorithm which gurantees the linear phase constraint, is developed that recursively produces the filter coefficient of linear phase FIR filter for a single block of data.

In this study, the motion compensating interpolation algorithm is presented. The presented algorithm allows the unblutted reconstruction of omitted frames. It is shown that the Walker & Rao's estimation algorithm using modified displaced frame difference combined with rectangulat adaptive measurement window increases the reliability of the estimation results. The remark ably improved image quality is achieved by change detection and segmentation.

The problem addressed in this paper is the accurate tracting of a dynamic target using outputs from a forward  looking infrared(FLIR) sensor as measurements. The important variations are 1) the spread of the target intensity pattern in the FLIR image plane, 2) target motion characteristics, and 3) the rms value and both spartial and temporal correlation of the back  ground noise. Based on this insights. design modifications and on  line adaptation copabilities are incorporated to enable this type of filter track highly maneuverable targets such as airtoair missiles, with spatially distributed and changing image intensity profiles, against, background clutter.

To track the manenvering target and to derive the Filter using state estimation and information in real time, we derive adaptive Kalman Filter which reinitialize the internal filter mode.

This paper presents a feasibility study related to the design of a linear multivariable compensator for a model of boilerturbine system. The nonlinear dynamics are linearized at a operating condition. At the operating point an LQG/LTR compensator is designed. Simulations are included to illustrate the usefulness of this linear multivariable control law.

A reliable Analytical Redundancy(AR) based Fault Detection Scheme(FDS) that can detect, discriminate sensor fault and process fault is presented. And a Fault Tolerant Control System ( FTCS ) with the FDS that performs original control objective without considerable loss of control performance in the face of sensor/process faults is constructed. These propositions are valuable in the sense that it resolves the well known sensitivity problem and that sensor/process faults can be detected, discriminated so that effects of any fault can be promptly accomodated by reconfiguring control system structure automatically.

Several sufficient condition which guarantee the stability of linear state delay systems are derived. And the delay margin which guarantees the stability of the delay systems are presented.

For feedback control of a linear dynamic system the optimum linear slace regulator (OLSR) can be implemented only if all state are available for feedback. This work demonstrates that when only the output state is available for feedback, a nonlinear controllers can be improved performance over that obtained by a proportional controller. This paper found the optimal control law by wellknown dynamic programming and principles of optimality. Thus, performance of both proportional and nonlinear controllers is compared with performance of optimum linear state regulator.

By using Lyapunov method, sufficient conditions for linear timevarying continuoustime and discretetime systems to be stable are presented under the assumption that the systems are slowly timevarying. Though it is not simple to find the stability regions immediately, one could find practical and large stability regions by constructing an appropriate algorithm.

This paper suggests mathemathical engine model which can simulate generalized gasoline engine. Suggested engine model contains air/fuel inlet element, intake manifold, combustion, engine dynamics. In order to analyze a gasoline engine, physical characteristics of engine and steady state engine data should be controlled. In adaptation for abrupt change of circumstance or drive conditions, this model can analyze important physical phenomena in the intake manifold by computer simulation. This model can also evaluate statuses of drive under various working conditions precisely. Therfore, this model suggests basic datum to evaluate the engine system which are needed in designing and development of engine controller.

A decentralised computational procedure is proposed for the optimal feedback gain matrix of largescale discretetime systems with timedelays. The constant feedback gain matrix is computed from the optimal state and input trajectries obtained hierarchically by the interaction prediction method. All the calculation in this approach are done offline. The resulting gains are optimal for all the initial conditions. The interaction prediction method is applied to timedelay largescale systems with general structures by extending the dimensions of coupling matices. A numerical exampie illustrates the algorithm.

In this paper, a new method for determining the order of AR, ARMA processes based on PLS (predictive least square) principle is proposed, This method using modified lattice algorithm which has additional step is amenable to online or adaptive operation and is more accurate than any other mpthod. Some computer simulations are presented to show the efficiency of the proposed algorithms.

In this paper we present a new method to control system design using a sliding mode. The objective of this research is to develop the algorithm for the sliding mode control which is able to remove the chattering phenomena and to reduce the reaching phase. To accomplish this approach we introduce switching dynamics instead of switching logics to obtain the sliding mode. Consequently, we can obtain the new design approach which is much simpler than the VSS theory. And there do not exist chattering phenomena in this method because the obtained control input are continuous. Simultaneously we can reduce the reaching phase by a suitable choice of design factor. Numerical examples are discussed as illustration.

This note considers the class of controllers with integral action which arise directly from appropriate system models. Via internal model principle approach, a corresponding class of selftuning controller is shown to have both integral action in controller and offset removal in the tuning algorithm. The key idea is to constrain the estimator in each step in order to ensure that dc gain of feedforward and feedback polynomial of adaptive controller are always equal, thus allowing the loop integrator to work properly.

This paper presents a robust discrete MRAC with the discrete e/sub 1/ modification. Discrete e/sub 1/ modification is proposed by extending the Narendra's e/sub 1/ modification to the discrete version. In this case the adaptive controller was reformed by Lyapunov's direct metbod. And the improved robustness by using this algorithm was showed by sector theory.

In this paper, a new design technique which uses weighted leastsqare approach for the solution of the pole assignment problem is represented. This technique maybe used to assign some closed loop poles to places which reduce the large system input and output variance due to near polezero condition. The leastsquare approach is also applied to the design of servo selftuning controller with integrator.

A robust adaptive control method for a stable nonminimum phase SISO system with unmodelled dynamics is proposed. The robust parameter estimation method of the system with bounded output noise and unmodelled dynamics is employed and a parallel structure is proposed to improve the robustness of adaptive control system. The local stability of the proposed system is shown. Computer simulations are done in order to compare the performance of the proposed structure with the basic structure on various circumstance.

In this study, the several control laws and switching planes are considered to eliminate the steady state error and to obtain the rapid response for variable structure control systems. An elimination algorithm is achieved by combination with each other control law, and the type of control law having the most rapid response is shown through some simulations.

One of the major reasons of Adaptive Control is to control time varying systems. In this paper new adaptive algorithms are suggested for a class of linear time varying systems that satisfy certain assumptions. These algorithms consist of three modules, modeling, parameter estimation and control. The key feature of this paper is that power series of time varying parameters are used for estimation.

In this paper, we introduce the supermicro computer based nonrealtime BTR type DNC system, which can provide valuable datas for production and management by monitoring the operation of NC equipments. (upto 48 NC machine) In accordance with manual job schedulinq, part program is combined with tool presetting datas and enables rapid downloading and uploading. An application to the production system of die & Mold plant is presented.

This paper represents the computer application techniques of initial and modification machining for dies with 3dimensional scluptured surfaces. All procedures from die design to die machining and measurement are covered. The component of modelling is data management and modification (extrapolation and smoothing), surface modelling, and nc program preparation. Also this paper introduces the utility for successful and efficient operation of system such as map generation, data communication, tool path verification, contour map generation, graphic processing of extrapolation and smoothing results, and CAD/CAM system interface. Examples are given to illustrate the modelling.

In this paper, NCTOP is an interactive NC auto programming system for twodimensional machining. NCTOP consists of seven modules, that is pattern form input, free form input, AutoCAD interface, logic processing, postprocessor processing, output, and database upgrade module. The geometry information of a workpiece to be machined is put into NCTOP program. Then it generates NC program which is final output. NCTOP uses the machining dates from the data base.

Despite of the low productivity, the manual programming for NC machining is still widly used because of it's economical reasons. In this study, the computer assisted manual programming system CAMP which assits the efficient verification of MCD(Machine Control Data) is presented. The system can detect sysntax errors, graphically display tool motions and eventually diagnose programming techniques. The case study is applied for the NC turning operations.

This paper describes tbe design and implementation of the adaptive controller to maintain the glood weld quality in gas metal arc welding process. The weld torch travel speed and the surface temperature are taken, respectively, as an input and an output of the welding control system. Because of the very complex phenomena of the process, the inputoutput dynamic model was experimentally identified by AIC (Akiake Information Criterion). Based on the model structure, the explicit model reference adaptive controller is simulated in order to regulate the output tempernture to the desired level.

For the automation of the manual soldering process through robot technology, two main tasks have to be achieved: Control of various soldering parameters and realization of flexible tool movements like human hands. In this paper a method for attaining these tasks is presented and analyzed.

This paper describes the application of the variable structure control(VSC) concept for the position control of electrohydraulic servomtor system. The basic philosopy of VSC is that the structure of the feedback control is altered as the state crosses discontinuity surfaces in the state surface with the result that certain desirable properties are achieved. The switching of the control function yields total(or selective) invariance to system parameter variations and disturbances, and closed loop eigen value placement in timevaring and uncertain systems.

Drawing pneumatic circuits by hand and searching for the error when the circuit is not properly constructed are very difficult. In this paper, a graphic simulation program for drawing and evaluating pneumatic circuit systems was developed. The porgram is menudriven style and pneumatic circuit can be easily drawn by selecting the pneumatic components from the menus. Simulation of the motion of each pneumatic component and testing of whether the circuit is constructed properly are possible with the software. This program was written in Turbo Pascal and also used the Turbo Graphix Toolbox. The system hardware requires IBM PC, XT, AT or compatibles, and Hercules Card.

In designing a high performance electrohydraulic control system for a large flexible structure, several flexible structural modes should be taken into account in a range of hydraulic control system bandwidth. The procedures of modeling a flexible mode control system and designing the high pass filter of load pressure feedback are presented. Example analysis varifies the presented analysis.

This paper considers modelling and control of ABS(Antiskid Brake System) which avoids dangerous wheel locking due to excessive brake pressure during the vehicle braking. The brake pressure is controlled by on and off's of solenoid valves via the variation of the wheel circumferential deceleration measured using tachosensors. The dynamic model between the brake pressure and the wheel acceleration of a vehicle is mathematically derived. The computer simulation shows that the threshold value of the onoff control is critical to the performance of the ABS.

The objective of this study is to develop a microcomputerbased adaptive controller for an electro hydraulic velocity control system subjected to the variation of system parameters. The step response performance of the system with the adaptive controller is investigated for the variation of the external load torque, the moment of inertia and the reference inputs, and compared with that obtained by PID controller whose gains are constant. The experimental results show that this proposed model reference adaptive controller is robust to the variation of system parameters and yield much better control performance compared with the conventionel PID controller.

In this paper, an iterative leaning control algorithm for the linear discrete system is proposed. Based upon the parameter estimation method, the learning for good tracking control is acqured through a sequence of repetitive operations. A series of simulation are performed to show the validity of this algorithm.

A digital seriesfeedback compensator algorithm for tracking timevarying signal is presented. The seriesfeedback compensator is composed of one closed loop pole / zero cancellation compensator and one desiredinput generator. This algorithm is applied to nonlinear hydraulic position control system. The hydraulic servo system is modelled as a second order linear model and cancellation compensator is modelled from it. The desired input generator is inserted to reduce modelling error. Digital computer simulation output using this control method is present and the usefulness of this control algorithm for nonlinear hydraulic system is verified.

Recently, in sequence control systems, high flexibility and maintenance of control software are required. This is because product life cycles become shorter and control specification must be changed frequently. The authors extend the concept of Safe Petri Net to develop the design and analysis tool for sequence control systems taking the safeness and notation of input/output functions into consideration. Extended Safe Petri Net (SNet) is proposed as such a new graph model and real time scanning algorithm based on SNet is developed.

This paper describes an application of the periodic scheduling policy to a class of flexible manufacturing systems. A heuristic algorithm is used to determine when each part should be loaded into the system. Simulation results based on a detailed model of metalworking manufacturing facility are presented.

In this study, a structured supervisory control for automatic assembly system is developed. And also diagnosis and faultrecovery capability of this supervisory control are discussed. This structured supervisory control is actually applied to Die Bonding Machine and is proved to be useful and to work well.

In this paper, the design and implementation of a multiprocessor based die bonder machine for the semiconductor will be described. This is a final research results carried out for two years from June, 1986 to July, 1988. The mechanical system consists of three subsystems such as bonding head module, wafer feeding module, and lead frame feeding module. The overall control system consists of the following three subsystems each of which employs a 16 bit microprocessor MC 68000 : (i) supervisory control system, (ii) visual recognition / inspection system and (iii) the display system. Specifically, the supervisory control system supervises the whole sequence of die bonder machine, performs a selfdiagnostics while it controls the bonding head module according to the prespecified bonding cycle. The vision system recognizes the die to inspect the die quality and deviation / orientation of a die with respect to a reference position, while it controls the wafer feeding module. Finally, the display system performs a character display, image display ans various error messages to communicate with operator. Lead frame feeding module is controlled by this subsystem. It is reported that the proposed control system were applied to an engineering sample and tested in realtime, and the results are sucessful as an engineering sample phase.

This paper presents a development of flexible bowl feeder without mechanical tooling for each specific part to be fed. This flexible bowl feeder is capable of recognising different orientations of components using optical fiber sensors and can be adjusted to feed different components by way of a simple programming method. The system would be suitable for feeding and sorting small components prior to auto,ated small batch assembly.

In this study, a model for the simulation of the material flow not only inside a robot cell with flexible handling sequence but also between robot cells is presented. A method for the connection of special simulation programs has been developed and a logic model between a real system and a simulation system is employed.

Althouah there have been many attempts to control weld quality in resistance spot welding processes, design method for an online feedback controller based upon process dynamics has not been suggested. This is due to the fact that the resistance spot welding is a highly complicated process, whice involves the interaction of electrical, thermal, mechanical and metallurgical phenomena. In this paper, an optimal control method based on FDM model with shunt effect is presented, which can regulate the nugget size, at the same time minimizing the control heat input. Optimal PI gain of the controller were determined by numerical optimization. Simulation results show that, as a result of the proposed optimal control, the weld nugget can be made to approach a desired nugget size with less control heat input than that required for the conventional spot welding process in the face of the shunt effect.

We propose a framework for modelling and operation management of robotic assembly cells via knowledge base. In the framework, each component of the cell is considered as a state variable, the relations among the state variables are stored in state transition maps(STMs) and then transformed into the form of knowledge. The assembly job tree(AJT) which includes the precedence relations and the constraints for assembly tasks is also described. Finally, an algorithm is presented to manage the cell operation.

An optical device, named "TMES", to measure the accuracy of printed scale of measuring tapes was developed. It has 11 optical sensors, that are exactly positioned every 100 mm on the 1 m frame. The sensors detect the beginning of the printed part and generate the puleses. The time difference between each signal pulse and the first sensor signal pulse is proportional to the error of printed scale. The measured data are trasmitted through a IEEE488 interface to a IBMAT computer and analyzed. The analysis software continuously displays the measured result on the CRT.n the CRT.

PCS(Process Computer System) is installed to monitor various kinds of process parameters in Power Plant and networked for synthetic monitor and event analysis in all site. But when an event is occured sequentially or simultaneously among the plants, it is difficult to analyze it because of different StardardTime in each plant. StandardTime Receiving Device is developed to solve this problem and development procedure is described here.

The implementation of MAP(Manufacturing Automation Protocol) NIU(network interface uint) is studied and presented in this paper. Hardwares and system softwares for network management, user interface, and medium access control are developed. Also LLC(Logical Link Control) protocol is implemented. Afterwards, The operations of NIU using selftest functions are tested. Compatibility between interfaces according to IEEE 802.4 standard is also tested.

This paper described network model system for FA with programmable devices such as PLC, CNC, Robot which basis on the MAP protocol and constructed with the broadband 1OMbps cables. Developed systems can comniunicate and control with programmable devices. Such systems we called are Remote PLC Control System, Remote CNC Control System, Remote Robot Control System.

Various types and huge volume of information such as process instructions, workin process and parametric data are created in a wafer fabrication process and should be provided to personnels inside or outside the facility. This article describes design criteria and functional description on the information system for smallscale wafer fabrication process to accomplish paperless fab and to support efficient fab management.

The architecture of a multiprocessor based programmable controller(MBPC) is presented. It consists of a host processor, processing elements, and Input/Output processors. Some problems in implementing such architecture are also described. To resolve them, we proposed and presented INFOBUS, a system bus for MBPC. The performances of INFOBUS and MBPC are analysed using both analytic models and simulations. Some results from the analysis will be given and validated. In case of 50% of BTI(Block Type Instruction) and 4 processors, the scanning time is shown to be 0.194msec/Kstep with some reasonable assumptions.

Recently PLC pursues faster scanning time, circuit confidence, reliability improvement, and smaller size. To obtain above all merit, custom IC(Gate Array) is developed. Custom IC includes 5 main blocks and 2 auxiliary blocks. The 5 main blocks process faster sequential instruction execution by only logic gate using hexa instruction code system. And the 2 auxiliary blocks generate baud rate clock (153.6 KHz, 76.8KHz) to communicate between PLC and computer or programmers.

Elevator system requires position and speed control at the same time recently. The control device of existing Elevator system making hardware is simplified by using microprocessor that have been developed. In this papers, it consists of contactless logic circuit using miroprocessor and digital components. This paper shows that as this system control voltage and frequency using PWM inverter at the same time, speed control is accurate, acceleration and deceleration is soft and passengers can be feel comfortably because speed change is a little during driving.

The purpose of this research is to develope a algorithm characterizing the grasp of the randomly fed objects using the tactile sensor. The tactile sensor used is composed of the 10 x 16 arrays of optical sensors and the planar resolution is 1.8 x 1.8 mm
$^{2}$ . The square and circular plate are used for the investigation of the characteristics of this sensor. The result shows that the measuring performance of the square plate is superior to that of the circular plate. Based upon this result the algorithm for the assembly of the electric plug was developed and was implemented using the minirobot. 
In this study, some issues on 3 dimentional object recognition and pose determination are discussed. The method employs a laser projector which projects a cyliderical light beam on the object plane where it produces a bright ring pattern. The picture is then taken by a T.V camera. The ring pattern is mathmetically the ellipse of which the geometrical parameters have the 3dimentional feature of the object plane. This paper gives the mathematical aspects of 3dimentional recognition method and shows experimentally the variations of ellipse parameters as the spatial deviation of the plane object.

Direct measurement of the relative position between the end effector of robot and moving objects reduces difficulties caused by the joint encoder reading and transformation. For those purpose, the online sensing method using PSD sensor was developed in this paper. The sensor was calibrated on the precision table. Then, the relative position of a moving objects on the conveyor was measured while the robot was tracking the one.

In this paper, we describe a novel robot sensor designed to measure some of the thermal properties of a gripped object. The thermal conductivity of the object can be calculated by the proposed method. And the thermal sensor response enables robot system to discriminate between objects made of different materials.

This paper deals with the rotationfree optical sensor which will be used for two dimensional robot tracking. This sensor consists of position sensing device and coordinate transformation unit which transform the sensor coordinate into base coordinate. A new coordinate transformation algorithm which use analog signal is presented, and an analog circuit based on this algorithm is constructed. To prove the rightness of this algorithm, same experiments are carried out. And the performance, of this sensor is investigated.

A PID controller must need not only good servo response but also little operation of a control valve. We suggest a direct poleplacement PID selftuning algorithm using the structure of derivativeofoutput controller and Bezout identity. This algorithm can much reduce the change of output of controller and well follow the desired trajectory.

In this study an oneboard micom controlled precision temperature control system has been developed. The digital temperature control system is consisted of an oneboard micom as digital controller, a 12bit A/D and D/A converter, a power amplifier, a NTC thermister, a preamplfier and a heat chamber. An operating control program for the control system was written in Z80 machine language. A DualPID predictor control algorithm was proposed. Experments were conducted with different sampling time and limitted error value. As a result, the temperature in a heat chamber can be well controlled within + 0.2 .deg.C when the sampling time was applied to 10 sec and the limitted error value + 0.5 .deg.C under the dualPID predictor control algorithm. By means of oneboard micom overall system has been reduced in size and volume, thus the system becomes compact and less expensive.

In this paper, it is described that digital backup control system is developed and applied to the thermal power boiler control system which consists of many kinds of analog electronic cards. Also, by the installation of this system in ulsan power pllant, it was proven to have excellent capabilities of fault detection and digital backup control.

A tunable reducedorder distillation model is proposed for realtime applications. To develop the model, a binary distillation column with MaCabeThiele assumptions was considered first and then the governing equations for the column were reduced to a simplified vector differential equations using the collocation method combined with cubic spline interpolation function. The final reducedorder model has four tuning parameters, relative volatilities and liquid holdups for rectifying and stripping sections, respectively. To assess the applicability of the developed model,the realtime adjustment of the model was tried by recursively updating the tuning parameters using the BKF algorithm. As a result, it was found that the reducedmodel follows the simulated distillation process very closely as the parameters are improved.

Pressure and flowrate control systems for a supercritical extraction process are designed and analyzed. To do this, the dynamic model was first set up using the performance equations of control valves, CO2 compressor and the state equation of the supercritical fluid. Using this model, optimum pairs of manipulated and controlled variable which give least steady stat interaction are determined though the relative gain analysis.

In this paper a pH process of a weak acid with a strong base is modeled into a bilinear form, and a selftuning pH control algorithm which is robust against initial values of solution and disturbances is presented. The control algorithm employs the recursive least square method for the parameter estimation and the generalised minimum variance criterion as the objective function. The computer simulation shows that the tracking of desired pH values is obtained in satisfactory manner regardless of the initial values chosen for the process.

A sequentialclustered integrator based on GEAR method is developed for the purpose of dynamic simulation of chemical processes. And a single simulator structure capable of employing various integration approaches is designed and its efficiency and flexiblity is evaluated. Sequential integration method is superior to simultaneous method for the process without recycle, but simultaneous method is very powerful for the coupled process with recycle.

In this work, we developed a realtime versatile advanced control and supervisory software for a personal computer control. This software, basically, has background and foreground tasks which are performed in parallel at real time. First, background tasks are composed of controls of various kinds, reports and inputouput of signals etc, which are performed every sampling time. Second, foreground tasks are observation of operation conditions, data search, regulation of controllers and graphical design and display of processes, which are performed by users request. Additionally, this software has the functions of transporting data and composing distributed control systems, and all background tasks are composed of combination of unit function blocks.

Usually the first three joint variables (major link) and the next three joint variables (minor link) are used to determine the position and the orientation, respectively, of 6 degreesoffreedom robot manipulators. In this paper, the Jacobians of 20 major links and 6 minor links are calculated to find the positional maneuverability matrices and orientational maneuverability matrices. Then the kinematic characteristics of the major and minor links are examined. Also we gave the measures of maneuverability and the controllability of the links for the figure of merits of robot manipulator design.

The multitask operating system called HRMTOS (HUNDAI Robot Multitask Operating System) was developed for concurrent execution. HRMTOS consists of condition interpreter, queue constructor, task scheduler. Condition interpreter checks the status and condition of request, queue constructor makes queue according to the checked result by condition interpreter, and task scheduler finds the task that will be urgently executed by priority of queue after pending the current excuting task. HRMTOS could execute teaching, playback, monitoring function of multirobot and could be used more effectively than other robot controllers.

This paper describes an experimental system for automatic robot programming, The SNUARPS (Seoul National University Automatic Robot Programming System). The SNUARPS generates executable robot programs for pick and place operation and some simple mechanical assembly tasks by menudriven dialog. It is intended to enable the user to concentrate on the overall operation sequence instead of the knowledge regarding the details of robot languages. To convert task specifications into manipulator motions, the SNUARPS uses an internal representation of the world. This representation initially consists of geometric database from CAD system and is updated at each operation step to reflect the state changes of the world.

The main subject of this paper is the development of new ROBOT CONTROLLER, which can support MULTITASKING and MULTIROBOT functions. The system consists of various kinds of CPU modules according to their independent jobs. Acceleration and Deceleration profile is given in order to achieve the smooth robot motion and high cycle time. Further the communication capacity should be upgraded to meet the various kinds of peripheral PA devices.

Movement order program of robot operating program is generally made by teachin method. Therefore in most cases it is sufficient as long as the robot system shows a reguired repeatability for the working conditions. But the trend in the robot application moves to the automatic generation of the working programs. A mathematical robot model similar to the reality is necessary for the analysis of the kinematic transformation of the robot system. The purposes of this paper are to make a better describing form and to suggest an automatic algorithm for kinematic parameter identification.

A Teaching and Operating Expert System (TOES) was designed in order to perform a task effectively which is inaccessible to man. Once an error occurs in the middle of the task operation, the automatic mode is converted into a manual mode. After recovering the error by the manual mode, the manual mode should be converted into the automatic mode. It was necessary to improve the manual mode in order to increase the availability of a manrobot system, a part of the human interface technique. Therefore, the Error Recovery Expert System must be constructed and developed.

Basic types of bilateral servo systems were described and practical consideration in the bilateral servo controller design was introduced. Power assistance to the operator is essential for high efficiency and accurate force reflection is necessary for dexterous manipulation. This paper shows a controller structure under development at KIMM which employs nonlinear friction compensation and memory based gravity compensation technique for efficiency and dexterity.

Recently the needs for industrial robot system of high speed continuous path motion such as sealing application have been increasing. To apply to this kinds of work, dynamic analysis for Hyundai 8601 robot has been carried out by experiment. For the good design of high performance robot manipulator, accurate analysis of dynamic characteristics is important especially for current semiclosed loop control type industrial robots.

This paper presents a method to balance the links of an articulated robot for an unknown payload using an automatic balancing mechanism. The balancing masses are controlled to move in their appropriate locations so that the joint torques of the links are eliminated. After balancing the mass of the payload is obtained from the balancing conditions. Based upon a series of simulation studies some results are discussed.

In this study a microcomputeraided design program has been developed to design and analysis for the digital adaptive control system. DACS(Digital Adaptive Control System) program has been written in GWBASIC language which is suitable for IBMPC compatible. The dynamics of each element was modulized and described by linear difference equations. By the aid of this program, sampling time, the number of bits of A/D and D/A converter and the stability for the digital adaptive control system can be determined. In order to estimate the system parameters an online identification and a regression analysis method are utilized. The simulation results have been well agreed with the experiments. To demonstrate the utility of this program, an adaptive control system has been designed for airheating system.

In design of the plastic injection mold, Almostdatas for the mold are decided During the assembly design. In this study, the designer will be able to carry out not only components design but also creation of machining data automatically, by the assembly data.

In this paper, the authors introduce the concepts of neural computer systems which have been studied over 25 years in other countries. And also we illustrate the models of neural networks suggested by researchers. Our fundamental hypothesis is that these models are applicable to the construction of artificial neural systems including neural computers. Therefore we assume that neural computer systems are abstract computer systems based on the computational properties of human brains and particularly well suited for problems in vision and language understanding.

Shrinkrings (Stressrings) are used in the fabrication of dies for cold forming and powder compaction processes to increase the allowable pressures for a given die material. Optimum procedures are to minimize a die thickness under the conditions that the stress distributions in the die and stressrings utilize fully the strength available in each of the die elements. This paper proposes a new approach, where the maximum allowable shrinking pressures are calculated on shrinkage plans in the radial direction and the fractional shrinking pressures below the maximum allowable pressures are used as the design values. Two criteria for the optimal die design are used: Maximum shear stress limit for onepiece dies and zero tensile stress limit for combined dies. A computer program, DIECOM, is developed for illustrating the computeraided design procedures. Finally, examples for each case are presented in this paper.

The purpose of this study is to develop a personal computer aided 3D geometric modeller. To perform this study, we set up a cube, a cylinder, and a prism as a primitive in the first segment of this study. By modelling the 3D object through their transformation, addition, and subtraction, we proved the validity of the developed algorithm and its computer program. Some examples show the results of applying the program to modelling a few simple shape of the machine parts. These results met the first aim of this study.

As an efficient way of modeling bodies of complicated shapes, the sweeping and skinning operations have been implemented. These two operations are very powerful modeling method when the body is defined by the cross sections at various locations. For the implementation, the data structure for storing the cross sections and the resulting three dimensional body has been constructed. The resulting object is defined by the boundary representation based on the nonuniform nonperiodic Bspline surface.

The objectives of this paper are to establish an equivalent magnetic circuit of the hybrid PM step motor and to apply the canonical piecewise linear eqation and the Kazenelson algorithm of solving the nonlinear magnetic circuit, so that the static holding torque and the magnetic properties of the motor are determined. The results of that are applied to the design of the hybrid step motor.

As the power facilities grow up, the optimal operation and the best maintenance of power plant can not be overestimated too much, which can enhance the plant availability and reliability much further. In this respect, fault diagnosis methodologies of dynamic system which is timevaring and strongly nonlinear have been studied. On of them is to use algorithm which is based on timeinvariant, linear system, but this is not so nice a method for applying to power Plant. Therefore, the study on other techniques using Artificial Intelligence (AI) is under way. In this paper, the existing ways of fault detection are surveyed and their problems are also discussed.

Tuning of digital STC parameters for the system with unknown and time varying parameters as well as delays presents many preconditions and difficulties. In order to eliminate these difficulties and to enhance STC control functions, the present study employs the metbod of expert control for STC and to implement it in nodminimum phase plants. In this paper, implimedtation of the Expert STC for process control, STC. parameters is carried out by tuning systematically and by applying these parameters to nonminimum phase plant. This controller supervises a status of the plant, and robust control function for time delay or change of the plant is demonstrated by computer simulations.

In this paper computer control system is introduced for the reheating furnace in steel making process, and in order to support computer system and process operation, the configuration of expert system application is demonstrated breefly, which is under development.

PCS(Process Computer System) is installed to monitor various kinds of process parameters in Power Plant and networked for synthetic monitor and event analysis in all site. But when an event is occured sequentially or simultaneously among the plants, it is difficult to analyze it because of different StandardTime in each Plant. StandardTime Receiving Device is developed to solve this problem and development procedure is descried here.

In this paper, we consider conventional and modified proportional navigation guidance(PNG) laws for a random maneuvering target. By means of Lyapunov function approach, we show that an ideal missile guided by the conventional PNG law can always intercept a random maneuvering target if some specified initial conditions are satisfied and the navigation constant is chosen sufficiently high. In addition, we propose a modified PNG law. At the final phase of pursuit, the proposed guidance law has a better acceleration profile than the conventional PNG law.

On this paper, the application of Eiganstructure assignment to flight control system design is presents. Both output feedback and constrained output feedback are considered. The computer implimentation of the algorithm is discussed including the utilization of real arithmetic for complex conjugate eigenvalue. And the example include a stability augementation system, an autopilot decoupled mode control.

Subsonic aerodynamic force and moment characteristics have been determined for parametric groups of launchers. Geometric variables considered in the experimental study were body length, nose planform, fin planform (bodyalone contained), number of stage and existence of extrnal booster, All data were obtained for angles of attack from 0 to 27 deg in 3deg. increments.

In this paper, an arbitrary mission of a spacecraft from launch to orbit injection is investigated for some launch trajectories. Launch sites are selected arbitrarily from various locations in Korea. For this study, a general purpose simulation program is developed. In this program, the Earth is assumed to be a rotating ellipsold. The launch vehicle is treated as a 3D, 6DOF rigid body. The developed program has been evaluated by calibrating it to the 3stage N1 rocket developed in Japan and launched at Tanegasima. The simulated trajectory from the firststage ignition to burnout of thirdstage main engine is compared with the previously tested N1 rocket trajectory in Japan. The result shows that the program works properly. Using the proven program, we have investigated possible launch locations in Korea, namely Marado, Sungsan and Kuryongpo. The problem of polar orbit injection is also examined.

GaAs solar cells may be the most attractive and efficient power source of a satellite. GaAs is more radiation tolerant and less temperature sensitive than widely used silicon.
$Al_{x}$ Ga$_{1x}$ As/GaAs solar cells have been designed and fabricated by Liquid Phase Epitaxial method. GaAs solar cells, of which structure is about 0.2 .mu.m p$^{+}$  window layer, 0.61.O .mu.m Gedoped player. 3.mu.m nGaAs layer and n$^{+}$  buffer layer, have been characterized as a function of operating temperature from 25 .deg.C to 130 .deg.C. Open circuit voltage decreases linearly with increasing temperature by 1.41.51 mV/ .deg.C while degradation of silicon solar cells is about 2.22.5 mV/ .deg.C, short circuit current does not increase much with increasing temperature. Relative efficiency decreases with increasing of temperature by about 0.210.29 %/ .deg.C. Efficiency degradation of silicon solar cells with temperature is known to be about 0.5%/ .deg.C and our results show GaAs solar cells may be an excellent candidate for concentrated solar cells.ells. 
Through the introduction and understanding of the total system design and the system integration for sounding rocket with the purpose of the pure scientific research, that is, atmospheric research, resources observation, space observation and etc, the system will be operated easily, readily and effectively in the development and manufacturing of launching vehicles for the scientific satellite and communication satellite in the future.

In this paper, a zerovelocity update technique to improve navigation accuracy of a SDINS(Strapdown Inertial Navigation System) has been studied. An indirect feedback Kalman filter which includes SDINS error equations based on a quaternion between bodyfixed frame and local level navigation frame is employed for processing zerovelocity updates in an onboard navigation filter. Simulation results for landmobile vehicle show that the zerovelocity update technique make a significant contribution to improving SDINS performance without any external aids.

In this paper, the problem of regulation in the presence of a known firing disturbance is considered. We show how one can apply a disturbanceutilizing control(DUC) theory to a actual gun servo model. Applied disturbanceutilizing control theory is established by combining LQ regulator and reduced order observer in the discrete time domain. To see the performance of the applied method, computer simulation results are given.

A strapdown inertial navigation system fabrication utilizing dynamically tuned gyroscope was finished as a first stage development. So it's design, fabrication and tests are reported. Although this system lacks in accuracy compared with the cimballed system, factors such as low cost, small size and lightness make it useable in wide range of applications. The initial cost for investment is relatively cheap, and so it is best suitable for local development in various kind of inertial navigation system. Since all of the locally used systems are imported and even with it's close relation to the military, foreign technical transfer is practically nonexistent. The independent local development of such system at a time of domestic initation in aerospace and defense industry, can be seen as a significant milestone in the advancement of the inertial navigaion system field.

The inertial measurement units in Strapdown System are characterized by the fact that sensors directly mounted to the vehicle frame. So the sensors are subjected to the translatory and rotation dynamics of the vehicle. The sensor outputs involve many error terms. We must compensate the error terms for accuracy improvement. The method which identify the error parameter is studied and suggested.

In the existing GFCS (Gun Fire Control Systems) there is sometimes the problem of the miss distance which is between a target and the projectiles from gun and cannot be neglected. This errors are difficult to reduce either in the gun design phase or by precalibration exercise. In this paper the CLSA (Closed Loop Spotting Algorithm) which is applied to improve the performance of the GFCS is porposed and analysed. The results simulated by Monte Carlo technics show us better performance than the existing GFCS.

The performance of a strapdown attitude reference system(SDARS) under dynamic environments was analyzed by means of computer simulation. The study is aimed toward the performance evaluation in the presence of translational or angular vibration during 20 sec of flight time. The simulation was based on the error model of rate gyro, and Euler angle algorithm was employed to compute the attitude.

A study on Automatic Meter Reading and Load Control System used to Distribution Line Carrer has started since 1986, and the field trial system for AMR and LCS was installed to KEPCO on June, 1988. Now this system will be operated for testing. This paper will describe project and planning for field trial of KEPCO.

SCADA (Supervisory Control And Data Acquisition) system for control and supervision of power system is now being operated in KEPCO with high performance. But the performance of RGS(Report Generating System) is comparatively insufficient because the SCADA system was produced for special purpose performing in real time. Therefore, we have performed this study applied for the purpose of processing the statistic data of power network and improving facilities and availabilities of the system.

Current substation SCADA system has only one control center where two or three operaters monitor and control more than 40 substations spreaded far and wide. The scope of this system is too wide to cope with the concurrent multiple accidents in electric network by the typhoon and large mow. This paper describes the method to organize the optimal structure of substation SCADA system using remote console and small scale control center.

Digitalization of processor and increase of control information need the fiber optic data network which has excellent noise immunity and highquality, widebaudwidth information transfer capability. It is aimed to offer a fiber optic dataway system for measuring and control the power plant, that has high reliability, high data transmission rate and small cable duct with large transmission capacity.

The active filter system for harmonic current compensation is presented in this paper. The active filter, composed of a threephase voltagetype PWM inverter and the capacitor, compensates both the harmonic currents and the reactive power by injecting the PWM current to the ac line. This paper describes the principle of harmonic current compensation, the calculation circuits for the harmonic currents to be injected, the several compensation characteristics. Also the experimental results are shown to verify the theory proposed in this paper.

This paper is investigated on the improvement for steam temperature control at Boryung coalfired drum boiler type thermal power plant. The steam temperatur control has been mainly operated by the feedback controllers. Automatic controllers are bounded and difficult. Because boiler system is nonlinear and the system time delay is very large. Optimal regulators including predictive feedforward and differentiate control are synthesized and some improved output results are presented.

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

We have investigated imaging performance of digital rediography system by measured several imaging properties such as sensitivity characteristic, scatter fraction, modulation transfer function and detective quentum efficiency. The results show that digital rediography system has linear sesitivity to incident energy, excellent scatter elliminaility and high detective quentum efficiency, but poor modulatio transfer function.

Using back electromotive force(EMF) signals of a brushless DC motor, the sensorless microprocessor controlled drive system was developed. In this new commutation method, the manual pulses are used for relatively short accelerating phase and then the exact commutational positions are detected based upon the back emf signals. The hardware and software implementations with the experiment to compare the performance of the developed system with the, conventional system using hall effect sensors are included. By reducing the number of the required sensors in the artificial heart control system, the total reliability will be incresed.

As a muscular contraction is sustained, the power spectrum of the myoelectric signal is compressed into lower frequencies. The median frequency appears to be the prefered parameters to monitor this compression. This paper describes a technique and a device which provide an estimate of the median frequency using a TMS32020 DSP chip and IBM PC for tracking of this parameter. Results obtained from myoelectric signal are presented and discussed.

In this paper, we develop 256*256*8 bit realtime image digitizer with the IBMPC as a image processor. With this developed system, we process a chest Xray image, which is a medical image, using the adaptive algorithm. In the future, we consider that the application of this developed system is extended in the various fields.

In this study, 8 ch. FECG signal storage system with general cassette recorder and amplifier is developed, and simulated LMS algorithm. In future we construct real time FECG monitor system that is used digital signal processor.

Image restoration technique using dual sensor is presented in this paper. Digital Radiography image (1024xlO24) is obtained by conventional resolution sensor. We also obtain local DR image data by high resolution sensor. Two dimensional maximum entropy power spectrum estimation (2D ME PSE) is applied to low resolution image and high resolution image for the purpose of the power spectrum estimation of each image. A class of linear algebraic restoration filter, parametric projection filter (PPF), is derived from the power spectrums of each image. It is shown that the noise energy may be considerably reduced through the PPF.