제어로봇시스템학회:학술대회논문집
Institute of Control, Robotics and Systems (ICROS)
 기타
Domain
 Machinery ＞ Robot/Automated Machinery
1990.10

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

In this paper, we propose a new manipulator control scheme based on the CMAG neural network. The proposed control consists of two components. The feedforward component is an output of trained CMAC neural network and the feedback component is a modified sliding mode control. The CMAC accepts the position, velocity and acceleration of manipulator as input and outputs two values for the controller : One is the nominal torque used for feedforward compensation(M1 network) and the other is the inertia matrix related information used for the feedback component(M2 network). Since the used control algorithm guarantees the robust trajectory tracking in spite of modeling errors, the CMAC mapping errors due to the memory limitation are little worth consideration.

In this paper we propose the force controller using a neural network for a redundant manipulator. Jacobian transpose matrix of a redundant manipulator constructed by a neural network is trained by using a feedback torque as an error signal. If the neural network is sufficiently trained well, the kinematic inaccuracy of a manipulator is automatically compensated. The effectiveness of the proposed controller is demonstrated by computer simulation using a threelink planar robot.

In this paper, the dynamic modeling and a tipposition controller of a singlelink flexible manipulator are developed. To design the controller of a flexible manipulator, at first, it is required to obtain the accurate dynamic model of manipulator describing both rigid motion and flexible vibration. For this purpose, FEM(Finite Element Method) and Lagrange approach are utilized to obtain the dynamic model. After obtaining the dynamic model of a singlelink manipulator, a controller which computes the input torque to perform the desired trajectory is developed using neural network.

In this paper ,ie propose a hybrid position/force controller of a robot manipulator using doublelayer neural network. Each layer is constructed from inverse dynamics and Jacobian transpose matrix, respectively. The weighting value of each neuron is trained by using a feedback force as an error signal. If the neural networks are sufficiently trained it does not require the feedbackloop with error signals. The effectiveness of the proposed hybrid position/force controller is demonstrated by computer simulation using a PUMA 560 manipulator.

Learning control of a robot manipulator is proposed using the backpropagation neural network. The learning controller is composed of both a linear feedback controller and a neural networkbased feedforward controller. The stability analysis of the learning controller is presented. Three energy functions are selected in teaching the neural network controller : 1/2.SIGMA.vertical bar torque error vertical bar
$^{2}$ , 1/2.SIGMA..alpha. vertical bar position error vertical bar$^{2}$ + .betha. vertical bar velocity error vertical bar$^{2}$ + .gamma. vertical bar acceleration error vertical bar$^{2}$ and learning methods are presented. Simulation results show that the learning controller which is learned to minimize the third energy function performs better than the others in tracking problems. Some properties of the learning controller are discussed with simulation results. 
Conventional back propagation learning is generally characterized by slow and rather inaccurate learning which makes it difficult to use in control applications. A new multilayer perception architecture and its learning algorithm is proposed that consists of a Kohonen front layer followed by a back propagation network. The Kohonen layer selects a subset of the hidden layer neurons for local tuning. This architecture has been tested on the inverse kinematic solution of robot manipulator while demonstrating its fast and accurate learning capabilities.

In the existing robot programming methods, offline method becames important role of programming because of improvement of hardware and software of PC. The purpose of this study is to develop practical robot programming system for polishing task using PC. In the first place, we have investigated the existing robot programming systems, and derived the requirement of this programming system from the existing systems. And we have decided the structure of this system. After that, we have developed this system. Using Windows software, this programming system has man/machine interface function. So users can use easily and quickly.

In the system of contemporary factory, the frequent generation of robot program reduces the efficiency of robot working. In this study, the SIOPS (Soldering and Insertion Offline Programming System) that automatically generates the robot program is presented. The system can change the parameter about soldering and insertion interactively and generate the robot tool path.

An algorithm is proposed to determine the optimal contact forces of robotic hands, where the soft finger contact as well as the frictional point contact are considered. Especially, the algorithm can be efficiently applied to the case of multipoint contact by innerlink as well as fingertip. To show the validities of the algorithm, several numerical exampies are presented by employing a robotics hand with three fingers each of which has four joints.

An optimal graspiftg algorithms for a multifingered robotic hand is proposed, where a new quality measure is developed to evaluate taskoriented as well as stability by modelling the tasks as ellipsoids. To show the validities of the proposed algorithm, several numerical examples are presented by employing a 3fingered robotic hand.

A number of performance measures have been proposed for the quantification of dexterity for kinematically redundant manipulators. The use of such measures is especially important for kinematically redundant manipulators since they can satisfy the subtasks such as singularity avoidance and obstacle avoidance in addition to satisfying a specification of endeffector motion. In this paper, the advantages and disadvantages of performance measures proposed up to date are compared through simulations under the same environment. Besides, a new dexterity index for manipulators with multiple degrees of redundancy is proposed and shown to be effective through the simulation.

This paper describes a method of calculating the viewing parameters of a camera using a perspective view of a circular pattern. The proposed method determines the angle of pan, tilt, and swing, as well as the distance from the camera to the reference point of a world coordinate system, using simple equations. The proposed method is so simple and accurate that when used in a wellcontroled environment as in robot vision systems or visual inspection systems it may even seem trivial.

As a preliminary study for the application of robots in electrical power company, several case studies in foreign countries, such as a substation patrol robot, a robot for nuclear power facilities, an insulator cleaning robot and a robot system for livedistribution line works, have been carried out for the application of robots. Various application fields, applicationrelated problems, technical difficulties, and the feasibility of applications are studied.

Analog systems could not fully deal with instrumentation and control facilities in power plants since these facilities have become more and more complex on account of various needs for large power generation capacity highly cost effective operation to conserve energy and operation management to prevent pollution. Digitalizations of control device and augmentations of control informations increased the necessity of fiber optic data networks which have excellent noise immunity and high quality wideband with information transfer capability. The goal of this study is to offer a fiber optic dataway system which has high reliabilities, high data transmission rates and small cable ducts with large transmission capacities for the instrumentation and control of power plants.

In this paper, the design of boiler drum level control system is important in power plant is studied. First of all, pressure compansation of level for boiler drum which is close tank with high temperature and pressure is designed. Physical penomena of drum level are analysed, controllers are designed, and simulation results are shown. Designed controller have a good performance compare with PI controller. Predictive controller of boiler drum level control system is proposed. It will be good system for boiler drum level control to reject the disturbance according to lead increase or decrease.

The long distance electric culvert employing power cable is using and increasing to improve reliability and Capacity of power supply in KEPCO. The purpose of this study in development of management system which used to security, maintenance, operation, and facility protection efficiently in electric culvert.

Dynamic modelling and controller design technique for constantfrequency series resonant converter with buck type preregulator are mainly described in this paper. An equivalent circuit model is derived and a state equation is developed from this model. To improve the dynamic performance, a negative feedback of inductor current is added to the proportional and integral control of output voltage. Furthermore, an optimization technique with prescribed eigenvalue region is applied to the determination of feedback gains. With the presented design method, much better dynamic performance can be obtained.

The problem of constructing an observer for use in the control of superheater temperature with desuperheater is considered. The distributed heat input into the superheater is usually not available for use in the observer, and hence is treated as a disturbance. The observer theory for systems with unknown inputs is exploited and applied to the problem. Approximation of the heat input utilizing the specific heat input distribution pattern is also considered.

It is necessary to develop graphic MMI (ManMachine Interface) for the maximum utilization of our installed ADS (Automated Distribution System) for which an operator types and executes string commands. Considering the restrictions and characteristics of the installed ADS and the data for modelled distribution lines, we calculated the time of periodic data acquisition for large number of points. And we presented the performance and characteristics of the graphic MMI system which will be developed and which will facilitate SCADA(Supervisory Control And Data Acquisition) functions.

In this paper, An expert system for electric furnace with time delay is proposed. The expert system uses Fuzzy control theory. The conventional controller uses AutoTuning control theory. From experiment, we can obtain that the response of expert system is superior to the response of the conventional controller. In this experiment, the expert controller is implemented with the IBM PC. The 8751 One chip processor controling the electric furnace is used.

The control process involving pure time delays presents a continuing challenge to the control system engineer. The nonlinear nature of the delay which can be introduced into the system make the use of conventional control algorithms a poor prospect. The Smith Predictor was developed to alleviate this problem. Unfortunately the quality of control achieved with the Smith Predictor is known to be sensitive to modelling errors. Only recently have researchers attempted to quantify the Smith Predictor controller's robustness to modelling errors. In several studies stability boundaries were plotted as functions of errors in parameters. But the research results address the question of performance of Smith Predictor controllers, In this paper, the Rule based Expert Systems for performance improvement of the Smith Predictor controller are developed.

This paper discusses the possibility of applying fuzzy logic controller in a microprocessorbased brushless DC servo motor controller, which requires faster and more accurate response compared with other industrial processes. Limitations of fuzzy logic controller are also described.

The fuzzy logic controller which can be applied to various industrial processes is quite often dependent on the heuristics of the experienced operator. The operator's knowledge is often uncertain. Therefore an incorrect control rule on the basis of the operator's information is a cause of bad performance of the system. This paper proposes a new selflearning fuzzy control method by the fuzzy system identification using the data pairs of input and output and arbitrary initial relation matrix. The position control of a DC servo motor model is simulated to verify the effectiveness of the proposed algorithm.

A backpropagation neural network based system for a fault diagnosis of a chemical process is developed. Training data are acquired from FCD(FaultConsequence Digraph) model. To improve the resolution of a diagnosis, the system is decomposed into 6 subsystems and the training data are composed of 0, 1 and intermediate values. The feasibility of this approach is tested through case studies in a real plant, a naphtha furnace, which has been used to develop a knowledge based expert system, OASYS (Operation Aiding expert SYStem).

In general, the control of robot arms falls into two board categories (position control and force control). The joint interpolated trajectory schemes generally interpolate the desired joint path by a class of polynomial functions and generate a sequence of time based control set points for the control of a manipulator from a initial location to its destination. A digital position controller was designed and adapted to the industrial balancing manipulator. And also, the joint interpolated trajectory using 3rd order polynomial was generated in this study. The IBM PC used as the main controller and the trajectory planner had enough runtime capabilities. The 8097BH microcontroller is an integral pan of the joint controller which directly controls an axis of motion. The PI servo control system to treat each joint of the robot arm as a independent joint servo mechanism had satisfying performance, and a sequence of timebased intermediate configurations of the manipulator hand showed good continuity and smoothness on position and velocity of the manipulator's joint coordinates along the trajectory.

On force control of robot, the transient response of the force is as important as the steady state value. We analyze the force for an 1 d.o.f. model. Based on the analysis, we finds out a desirable condition of the control system parameters for stability of the force. We propose a force rate feedback control for implementation. Through experiments we shows that the force can be controlled stably for an arbitrary environment.

In this paper, the problem of designing a feedback controller for a robotic manipulator, which is activated by a D.C. motor through a gear train and a flexible shaft or chain, is considered. When the response of the closed loop control system is relatively slow, a satisfactory controller may be designed as a PID controller. As the speed of the control system increases, however, the spring effect of the linkage becomes profound, and as a result, the transient response exhibits a substantial oscillation. To eliminate this oscillation, it is necessary to design the controller based on at least a fourth order system model. This, in turn, requires the feedback of the entire state variables. In practice, however, only the position of the manipulator and the velocity of the motor are readily measurable. The state variable reconstruction method or a state observer cannot be used because of the system nonlinearities such as the Coulomb frictions. In this study, an alternative controller, which is based on delayed feedback of the output variable only, is proposed, and a successful delayed feedback controller is designed and implemented on an actual experimental manipulator.

In this paper, simulation results of a robust digital tracking controller on a robotic manipulator are presented. The objective is to follow a ramp reference input with zero steady state error in the presence of a disturbance and system parameter variations. Some of the difficulties are caused by the Coulomb frictions, the disturbance due to the gravitational pull, the spring effect of a link between the drive motor and the manipulator arm. Another difficulty is that, because of the nondifferentiable Coulomb friction, the digital control system cannot be represented as a discrete system. It is thus necessary to design the controller based on a discretecontinuous hybrid model. The controller is based on feeding back the state variables and augmenting the system by addition discrete integrators. The feedback gain parameters are obtained by applying the quadratic optimal control theory and then choosing the new weighting matrices to eliminate the limit cycle by using the describing function method for hybrid system.

This paper presents the new VSMFC (variable structure model following control) algorithm of a system with parameter uncertainty. In contrast to the conventional approach, continuous control inputs am used instead of discontinous ones and modified reachability condition for sliding mode reduces a reaching phase and a newly proposed sliding surfaces are used for the robust control of uncertain system.

In this paper, Proposed the robust controller for robot manipulator plus actuator with dynamic parameter uncertainties. In general, errors and uncertainties system parameters exist more or less between the actual system and mathematical model. To reduce these trems, used Lyapunov stability theorem. The performance of the controller is evaluated for the three degree of freedom robot manipulator plus actuator model with uncertainties of parameters and model errors.

This paper describes the DNC technology as a partial approach to a achieving dr system integration of NC machines. The study deals with development of a system to unify NC machines by utilizing a DNC. The system is provided to have the potential of offering ft environment and application fundamentals to construct factory automation.

The information with regard to the working range of lathe, cutting tool, cutting condition is managed as Database system for turning operation as one part of CAM system. Data with regard to the working range of lathe, cutting tool, cutting condition are stored by the DBMS(Data Base Management System) and can be added, modified, deleted and retrieved for realtime usages. Data stored in Database system are searched to select the most proper cutting tool and cutting condition with the input data fed from the design stage. Codes in regards to tool shape are displayed on graphic mode for easy selection for user and thus presents a good decision support for tool selection. The system developed in this work is operated by the pull down menu on the IBM PC/AT personal computer, or compatible series.

Rotationalfreesurface (RFS) is a special type of free surface whose two boundaries coincide. For NC machining of PFS, a rotational axis as well as three Cartesian axes are required. In this paper, we develop a fouraxis CAM module consisting of: a) Geometric modeling of RFS, b) CLdata generation, and c) Graphic simulation of machining operation. To test the validity and effectiveness of the developed module, several test cuts are made with Bridgeport CNC milling machine and compared with the graphic simulation.

In design of the plastic injection mold, most of drawings are composed of basic entities. It is very easy to produce many kinds of drawings by Group Technology. Group Technology is a technique In which part similarities are used to classify parts into part families according to either geometric shape and size or processing requirements. Almost data for the mold are decided during the assembly design. A system which shows a good interfaces between the design stage and producing part exploding Is developed using AutoCAD system and data conversion technique.

The automization of manufacturing lines may be accomplished by replacing the human operator with computer system. This paper describes an idea to fully automize the razor qrinding process. Now, in this system, to control the process, human operator must estimate the qrinded states and control the grinding machine continuously. We propose two methods to automize this process by using CMAC memory. One is about learning expertrules without direct communication with operator. And the other is complete selflearning method based on CMAC's learning algorithm. These ideas may be applied for another manufacturing processes.

Die finishing (polishing and lapping) after NC machining is characterized as one bottleneck process for reducing lead time. For automation of this typical manual work, a flexible polishing tool system using industrial robot has been developed. This tool system has three principal functions in order to achieve reduction of waviness, 3 D.O.F. compliance and constant pressure structure. This polishing tool shows that adaptability to free form surface is increased and programmability to various areas of die surface is also acquired.

Because of being operated to accomodate the load variation, fossile power plants need modern controller to satisfy frequency regulation, fast response and efficiency requirement. But in the way of developing the Digital Instrumentation and Control System that can adopt advanced control algorithms, power plant simulator must be available for safe and convenient test. In this paper, a simple drumtype boiler model is proposed, using Signal Flow Diagram that describes the system by a collection of basic time response representing the relationship between process variables. After verifying the model and thereafter building simulator based on the proposed model, we test the operation of controller and approve stability or efficiency of implemented control algorithms.

A multiloop programmable controller using graphic control language and a configuration graphic editor for designing control algorithm in graphic windows are developed. Using the graphic control language, large and complex control algorithm can be designed easily and the modification of control algorithm is simple. The proposed controller using graphic language can be effectively used for complex plants like power plants.

The digital instrumentation and control system for the large scale system like the power plant must have the form of the heirachical structure. Because most large scale system have many control and process signals and it is distributed in the vade region, it is necessary to partition them into several subsystems. Therefore, the role of SCS(Supervisory Control System) having the functions of controlling and monitoring for the status of subsystems is very important. In this paper, new SCS for the effective control of the large scale system is proposed.

The activated sludge process is a commonly used method for treating sewage and waste waters. The process is characterized by a lack of measurement instrumentations and control goals that are neither well defined nor well understood. In the present study the concept of fuzzy control is employed for such process in which a design method for fuzzy controller based on a multivariable fuzzy reasoning algorithms is investigated and then simulation results are presented.

Neural networks are a class of systems that have many simple processors (neurons) which are highly interconnected. The function of each neuron is simple, and the behavior is determined predominately by the set of interconnections. Thus, a neural network is a special form of parallel computer. Although a major impetus for using neural networks is that they may be able to "learn" the solution to the problem that they are to solve, we argue that another, perhaps even stronger, impetus is that they provide a framework for designing massively parallel machines. The highly interconnected architecture of switching networks suggests similarities to neural networks. Here, we present two switching applications in which neural networks can solve the problems efficiently. We also show that a computational advantage can be gained by using nonuniform time delays in the network.e network.

A motor group control system is developed for continuous manufacturing processes such as rolling process or electrolytic tinning process. The control system consists of four subsystems ; MultiFunction Controller (MFC), Flexible Motor Drive (FMD), Bulky Input/Output (BIO), Graphic Console and Simulator (GCS). A graphic control language, called Function Block Language, is used to configure the control algorithms for each subsystem. All subsystem are linked together thru a field bus to communicate data with each other.

The major difficulty in distillation column control lies in executing the set point tracking and the disturbance rejection, because of continuous changes in model order and dead time. For that, generalized predictive control(GPC) was applied to distillation column control. Recursive least square method was used to adjust the changes of model order and dead time. Quadratic progamming(QP) was used to solve the constraint problems in control action and the rate of control action. As a result of the simulation on the dynamic simulator(SPEEDUP) and the experiment on pilot plant, the ability of the set point tracking and the disturbance rejection was acceptable to apply to the real distillation column.

A new online optimization technique of a distillation column is presented. The equation decoupling method and the simultaneous solution method are combined to simulate the distillation process. The storage requirement is small enough to run on a PC. A improved successive quadratic programming is used to find the optimum operating conditions. The optimizer is intensively tested by using the dynamic simulator, SPEEDUP. After this, the technique is applied to a binary distillation column that treats methanol and water.

A neural controller for process control was proposed that combines a simple feedback controller with a neural network. This control was applied to distillation control. The feedback error learning technique was used for online learning. Important characteristics on neural controller were analyzed. The proposed neural controller can cope well with strong interactions, significant time delays, sudden changes in process dynamics without any prior knowledge of the process. It was shown that the neural controller has good features such as fault tolerance, interpolation effect and random learning capability

A online optimization scheme based on model in a binary distillation process is proposed. A reducedorder model utilized the concept of collocation is used as a process model and the recursive prediction error method is employed to identify the reducedorder model. The concentrations of end products are controlled by nonlinear adaptive predictive control algorithm. The objective function is constructed to find optimum operate condition for saving utility cost. The proposed optimization is scheme is tested through simulation studies in 13staged watermethanol distillation column.

A fuzzy control scheme has been proposed for a supercritical extraction process which has attracted much attention recently as a new separation technology. Based on the manual operation experience, three control pairs between manipulated and output variables are selected first and then seven membership functions are defined for control error and time rate of the error, respectively for each control pair, resulting in forty nine Fuzzy control rules. In addition to these, the membership functions are defined in two steps (coarse and fine) to enhance control performance. Fuzzy inference is performed using MAXMTN composition rule and defuzzified control output is calculated based on center of gravity method. The prosed Fuzzy control scheme has been assessed through numerical simulation. As a result, the proposed scheme shows good control performance comparable with that by INA(inverse nyquist array) which usually requires complicated design procedure.

For the position control of BLDC motors, 2.phi.3.phi. conversion unit, hysteresis band device, and PWM pulse generation are required. By utilizing a singlechip processor INTEL 8097BH, we cover all required functions. As a result, no analogue device is employed in this control scheme, so that it makes BLDC motor controller simple and flexible. Furthermore, the singlechip processor is utilized for the nonlinear coordinate transformation which is necessary for the linearization of the motor system.

Several optimum control algorithms have been proposed to minimize the robot cycle time by velocity scheduling. Most of these algorithms assume that the dynamic and kinematic characteristics of a manipulator are fixed. This paper presents the study of a minimumtime optimum control for robotic manipulators considering parameter changes. A complete set of solutions for parameter identification of the robot dynamics has been developed. The minimumtime control algorithm has been revised to be updated using estimated parameters from measurements.

This paper deals with the manipulator with actuator described by equation D over bar(q)
$q^{...}$ = up over bar (q,$q^{.}$ ,$q^{..}$ ) with a control input u. We imploy a simple method of control design which bas two stages. First, a global linearization is performed to yield a decoupled controllable linear system. Then a controller is designed for this linear system. We provide a rigorous analysis Of the effect of uncertain dynamics, which we study using robustness results In time domain based on a Lyapunav equation and the total stability theorem. I)sing this approach we simulate the performance of controller about a robotic manipulator with actuator.tor.r. 
Usually, robot manipulators in production lines are operated with reperting work trajectories. This paper presents the repeated adaptive learning algorithm for robot manipulates for the case of a trajectory. This algorithm uses the nonlinear dynamic model including the repeated friction compensating term, The advantage of the scheme is that It allows friction compensation which may be otherwise difficult for differently constructed models. A secondary advantage of the sheme is that it can also adapt to torque calculation in order to reduce the computational load of the control computer. To show the efficiency of the proposed controller, a computer simulation is performed for the planar robot manipulator with a 2 degree of freedom.

A neural network method is applied for recognizing the orientation o f individual parts being fed from a bowl feeder. The system is designed in such a way that a part can be discriminated and sorting according to every possible stable orientation without implementing any a mechanical tooling. The operation of the bowl feeder is based on a 2D image obtained from an array of fiber optic sensor located on the feeder track. The acquired binary image of a moving and vibrating part is used as input to a neural network which, in turn, determines t he orientation of the part. The main task of the neural network, here is to synthesize the appropriate internal discriminant functions for the part orientation using the part features. A series of the experiments reveals several promising points on performance. Since the operation of the feeder is highly programmable, it is well suited for feeding and sorting small parts prior to small batch assembly work.

Our N.C. sewing machine is a device that acts sewing automatically in reference to N.C. data made by data input device. Our N.C. sewing mechine is composed of induction motor that has electric clutch and brake, step motor that drives xy table. Sewing area of our machine is 170 mm * 100 mm, maximum sewing speed is about 2000 spm, maximum stitch length 6.2 mm. In actual sewing, synchronization between motion of needle updown and motion of xy table is a critical factor. In this paper, technology about synchronization will be profoundly discussed.

We studied about the adopted system to POSCO facility which is very complexity and various as an example for studying a facility diagnosis and management system that is starting to be distinguished in the nowadays. At the first, explain what's the background why POSCO adopted the system, and it's goal, and it's promoting direction and ways. And there after studied it's results and effective analysis. Finally inquiry about the future improvement and development items through problems which become known during the project periods.

An antiswing controller for an overhead crane in the stop position is designed. The developed antiswing controller improves on the poor damping characteristics of overhead crane by feeding back the crane acceleration as a function of swing angular speed. The experimental results show that this crane using the proposed controller yields small stop position error and rapid damping response characteristics.

This paper presents technologies to improve the control of an impact wrench. Impact wrench is a tool which is held by the electromechanical manipulator and used to fasten and loosen the bolts for remote maintenance of equipment in hostile environment. Vision system was developed to measure the distance and improve the positioning of the impact wrench. The vision system used two laser beams with a CCTV camera. Also, a torque adjusting method was developed to limit the fastening torque.

An automatic visual classification system is introduced which provides for measuring the length and diameter of coilform cores and dividing them into 5 different classes in terms of how far their length be from the desired length. This task is fully automated by controlling two STEP motors and by using image processing techniques. The classification procedure is broken into three logical parts. Fist, cores in the form of randomly stacked bundle are lined up one by one so as to be well captured by a cameras. The second part involves capturing core image. Then, it enters the measuring process. Finally, this machine would retain all tire information relating to the length. According to the final result, cores are sent to the corresponding bin. This considerably simplifies the selecting task and facilitates a greatly improved reliability in precision. The average classifying capability about 2 pieces per second.

In this paper, we propose a simple method to control the servo system with sliding mode in the parameter variation and disturbances. We show the comparison between the conventional sliding line and the new sliding line and the proposed sliding mode control. The performance of the fast response and no overshoot by using the proposed sliding line is obtained.

This paper proposes the design of the decentralized adaptive controllers which are an arbitrary interconnection of subsystems with unknown parameters, nonlinear ities and bounded disturbances. In order to exponentially converge the state and parameter errors, robust decentralized adaptive controllers are developed for stabilization and tracking the parameters. In the simulation studies of the decentralized adaptive control of a twoarea interconnected power system, the effectiveness of the proposed adaptive schemes is demonstrated.

This paper proposes a performance improvement to a control system with optimal state feedback control. In this paper, a simple and direct design procedure is proposed to design the model following controller. The scheme is implemented in a 16 bit micromputer with math corprocessor. Tests on a DC machine have been conducted.

The solution of problems in the automation of mold manufacturing process are high speed and resolution. To catch two problems at once, digital servo system will be essential replacing the analog servo system. In this paper, we tried some hardware/software approach including, designing basic structure, modeling of servo actuator, and designing & simulating of velocity & position controller. Finally, we constructed the hardware for 1 axis digital servo controller of NC system. As a result we obtained the design technics of digital servo systems for high speed NC system.

Gunner primary sight stabilization system is a fully integrated sensor package designed to provide the stabilized LineofSight. In this study, to improve disturbance rejection capabilities, two types of compensator (LQG/LTR, LeadLag) were designed and then stabilization performances were compared under severe offroad environment. Simulation results shows that the stabilization performances using LQG/LTR methodology is better than LeadLag methodology in spite of dynamic uncertainties.

Most verification of improvements for adaptive control schemes are. dependent on computer simulations, but these computer simulations have much limitation, because (if complex actual conditions of system. This paper is concerned with the constructions of a thermal process system for experiments with various control schemes. This thermal process system is composed of a water tank, PCXT, AD/DA converters power supply and thermal sensors. We estimate. the algorithms of poleassignment adaptive control in the manifold disturbances and environments, changing system dynamics. The system equations for thermal press are included.

This paper describes a neuralnetworkbased methodology for providing a potential solution in the area of process fault diagnosis. The existing neural network for fault diagnosis learn fault node by using pairs of singlesymptomsinglecause only. But in real plants, the effect of a fault propagates continuously from it's origin; different sensor values reflect this. In this paper, we suggest a new method which can handle the effect of symptom propagation. The proposed method can find the exact origin of the fault of which the symptom is propagated continuously with time.

Under the assumption that process input/output data are sufficiently rich to allow reasonable plant identification, a longrange predictive control method for SISO bilinear plant is derived. In order to ensure offsetfree behaviour of the control method, a new bilinear CARIMA model with variable deadtime is introduced. Furthermore, to extend the maximum output prediction horizon, the future predicted outputs in the bilinear term are assumed to be equal to the known future setpoints. With a classical recursive adaptation algorithm, the proposed control scheme is capable of stable control of bilinear plants with variable parameters, with variable deadtime, and with a model order which changes instantaneously. Several simulation results demonstrate the characteristics of the proposed bilinear model predictive control method.

A pressure tracking of hydroforming process is considered in this paper. To account for nonlinearities and uncertainty of the process. A fuzzy SOC based iterative learning control algorithm is proposed. A series of experimentals were performed for the pressure tracking control of the process. The experimental results show that regardless of inherent nonlinearties and uncertainties associated with hydraulic system. A good pressure tracking control performance is obtained using the proposed fuzzy learning control algorithm.

This paper presents a new self organizing fuzzy linguistic control (SOFLC) strategy for application to an arc welding process control. The proposed SOFLC is based on online modification of the control rules according to the extent of deviation of the one step ahead predictive output of the process from the desired output. The Predictive output of the process is estimated by a fuzzy predictor which is updated from the input and output data of the process. The rule base of the fuzzy subsets describing the control rules is modified by the improving mechanism based on the hill climbing approach. Simulation results show that this proposed SOFLC improves the response of the process in presence of the variation of the process dynamic characteristics.

The resistance spot welding process has been extensively used for joining of sheet metals, which are subject to variation of many process variables. Many qualitative analyses of sampled process variables have been successfully attempted to achieve a uniform nugget size. In this paper, the electrode movement signal which is a good indicative of the nugget size was examined by introducing a mathematical model with four parameters. A neural network method was applied for the estimation of the nugget size by four parameters. The prediction by the neural network is in good agreement with the actual nugget size. The results are quite promising in that the qualitative estimation of the invisible nugget size can be achieved without destructive testing of the welds.

In this paper, a realtime fault tolerant control system has been designed for the cooling system of the spent fuel pool storage. The fault tolerant control system consists of the fault detection part, the redundant actuator part(main and backup pumps) and the controller implemented on programmable. logic controller. This paper considers only the actuator fault whose detection is accomplished using Friedland's separated bias estimation method. This paper also shows the realtime experimental results from which it can be concluded that the designed fault tolerant control system exhibits satisfactory performance.

This paper represents a control method of improving the performance of direct drive robot. The direct transfer of torque and rotational speed of direct drive motor to the robot body without reduction gear makes the robot speed fast. However, the variation of inertia matrix and low friction cause the control difficult, and one more effort must be in the reducing the acceleration and deceleration time to reduce the cycle time. To fasten the cycle time and to improve the robustness of robot, one control method is developed, and implemented in the Goldstar DD robot. This method does not need to change the conventional PI type control structure, but one additional compensational control law is required. The control law can be obtained via inverse dynamic model of robot, and inverse model of existing control loop. The effects of this control law are shown in this paper.

In this paper, the stability analysis for the bilateral control of telemanipulator is considered. Twoport network model is used to describe the telemanipulator system. The stable conditions are derived using impedance matrix and passive network theory for two different types of bilateral control scheme. It is also shown that such conditions can be succefully applied to the nd.o.f telemanipulator system in which the kinematics and dynamics of master and slave manipulators are different.

In this paper, the dynamic modeling and tip position of rotating Timoshenko beam analyzed by means of FEM (finite element method) and Hyperstability MRAC(model referenced adaptive control) technique of each other. The governing equations of the rotating beams are drived from Hamilton's principle. The dynamic model of this multilink is drived by Lagrange approach. The shear deformation and rotary inertia are incorporated into a finite element model for determining the bending frequencies of the rotating beam. Simulation results for uniform cantilever beams by using the MRAC are compared with the available results. It will be shown that the proposed method offers an accurate and effective one to solve the free vibration problems of rotating beams' stability.

A robot manipulator and an obstacle are described mathematically in joint space, with the mathematical representation for the collision between the robot manipulator and the obstacle. Using these descriptions, the robot motion planning problem is formulated which can be used to avoide a time varying obstacle. To solve the problem, the constraints on motion planning are discretized in joint space. An analytical method is proposed for planning the motion in joint space from a given starting point to the goal point. It is found that solving the inverse kinematics problem is not necessary to get the control input to the joint motion controller for collision avoidance.

We develop an automatic trajectory planning system (ATPS) for painting robots by proposing a new trajectory planning scheme. The new scheme considers geometric modeling, painting mechanics, and robot dynamics to output an optimal trajectory (in the sense of coating thickness and painting time) based on the CAD data describing the shape of objects, The new scheme is implemented in SUN/4 workstation to develop an ATPS for painting robots. To test the validity of the new scheme and to illustrate the developed system, numerous runs are performed and analyzed.

This paper deals with the design technique of the controller for the discrete system control using Extended Safe Petri Net which is deduced from Petri Net as its subclass with a specific constraint and which is introduced to develop the design and analysis for the discrete systems. First, we propose a construction matrix suitable for the discrete systems that represent the marking flows which are the dynamic behavior of the discrete systems. Next, we develop a method that can design the controller for the discrete system control by analyzing the proposed construction matrix into the incidence matrix of Extended Safe Petri Net. Finally, the validity of the proposed method is shown by using the incidence matrix and matrix equation of Extended Safe Petri Net model.

This paper presents an approximate analysis of the serial production lines using decomposition technique. A serial production line consists of a series of unreliable machines separated by finite buffers. The serial production line is evaluated by approximation method, the results of which are compared with those examined by the discrete time event simulation, based on this approximation method, a gradient technique is proposed, which improves the efficiency of an operation of production line through the reallocation of buffers.

A modelling process of a robotic assembly cell and a method for analysis of the assembly cell operation through simulation are presented. An assembly cell including industrial robots is the subject of the model. The states of the assembly cell elements are taken as the state variables and the relationships between the states are described mathematically using the operators. An algorithm for the cell operation is developed from the relationships between the states and the information on the assembly task, and efficient analyses are performed by the simulation results.

For the robot manipulator in performing precision task, it is indispensable that the robot utilize the various sensors for intelligence. This paper presents the development and implementation of an integrated control system for the control of robotic manipulator, a feeder, a conveyor belt system, force/torque sensor system, and a photo sensor system. Micro controller board was constructed for hierarchical control of the system. To set up the program interactively, a user can make use of the software which includes the fulldown menu and a dialog box. The user can make progress the program quickly and easily by a mouse. The related software was written in C and assembly languages.

A mixed IP formulation is presented which concurrently routes and sequences the tasks on the MRS, reflecting the flexible characteristics. As a preliminary work for the performance evaluation of the MRS, a two robot system working on independent tasks is studied. Models for three types of the system capable of simple error recovery function are established using queueing model, and performances are evaluated and analysed.

Given the polynomial in z, P
$_{0}$ (z) = z$^{n}$ + a$_{1}$ z$^{n1}$ + a$_{2}$ z$^{n2}$ + ... + a$_{n1}$ z + a$_{0}$ , it is of interest to know how much coefficient a$_{I}$ can be perturbed while simultaneously preserving the stable property of the polynomials. In this paper, we derive the maximal intervals, centered about the nominal values of the coefficients, having the following property: the polynomial remains stable for all variations within these intervals. And then, under the unfixed weighted perturbation evaluate upper and lower allowable perturbations.tions.s. 
A new robust stabilizing method of uncertain system is proposed. The model uncertainty is considered to be the system matrix perturbations. The region of perturbed system eigenvalues are estimated by union of the disks which have the constant radius. Full state feedback control matrix which satisfies the new stabilization condition can be obtained by weighted LQ regulator or pole assignment technique.

All models of dynamical systems invariably have some measure of uncertainties associated with some of their dynamics. The recent approaches to establish robustness of stabilizing feedback control against the possible uncertainties have a serious limitation, that is their applicability only to the systems that satisfy the matching conditions. Such conditions are rarely met in general applications. If a particular system satisfies the matching conditions, the addition of an actuator will destroy the satisfaction of such conditions. In this paper, we develop robust control algorithm for uncertain multivariable systems in which the matching conditions are not necessarily met. We empoly Lyapunov's second method to derive robust stabilizing controllers which guarantee asymptotic stability against prescribed uncertainties. The derivation consists of transforming the original uncertain system to controllable canonical form and constructing a constant switching surface by designing the closedloop characteristics as a function of the uncertainties. Numerical examples are discussed as illustrations.

In this paper, we present a method of analysing perturbed linear system by pole sensitivity defined by the rate of pole movement with respect of perturbation. Pole sensitivity give us not only the rate of pole movement but also the directional information of the pole movement. We present a method of design of a LQR by considering the pole sensitivity and show that the suggested method guarantee the stability robustness of parameter perturbation.

We treat the problem of the robust tracking of a class of nonlinear systems which can be asymptotically decoupled in approximate sense by state variable feedback. A nonlinear control law is derived such that the tracking error in the closed loop system is uniformly bounded and tends to a certain small neighborhood of the origin. Simulation results show that simultaneous lateral and longitudinal maneuvers in airplane can be accurately performed in spite of uncertainty in stability derivatives.

Lately, at an epock of fullscale satellite ranching plan of Korea, T.T.C (Tracking, Telemetery & Command) is a indispensable part. In this paper, particular attention is given to orbit determination problem of role of T.T.C. Orbit determination, which is applied to Kalman Filter and Smoothing Filter, use the observation data which is given by satellite tracking radar system, and then the simulation is accomplished. As a result, it shows effectiveness.

A VSC law is derived for the attitude control of an orbiting spacecraft in the presence of disturbance and parameters variation using reaction jets. The switching surface was chosen to be a linear function of tracking error, its derivative and integral. Simulation results are presented to show that in the closedloop system, precise attitude control is accomplished in spite of uncertainty in the system.

The implementation of modern guidance law derived from optimal control theory requires accurate current states of target, for example, position, velocity and acceleration etc. But there is no sensors that measure the target states directly. So they are estimated from measurable data. For atmospheric missile engagement, direct application of the modern guidance laws may result In deterioration of Intercept performance because of poor observability associated with angles onlymeasurements by passive seeker and homing geometry. In this paper, a trajectory modulation method called "adaptive Intermittent maneuvers" is added to the modern guidance law, so the observability is enhanced and, consequently, improved the intercept performance. The estimation algorithm called "modified gain pseudomeasurement filter" is used for tracking filter. It is assumed that the passive seeker measure the angles between line of sight and Inertial frame. The MonteCarlo simulation for realistic airtoair Intercept scenario are conducted to demonstrate the effectiveness of intermittent maneuvers.ermittent maneuvers.

A design of aeroload simulator is proposed which is used for the dynamic test of guided missile servoactuation system. Since the simulator is linked to the servoactuation system, it is a twoinput oneoutput system. The mathematical model of the aeroload simulator is derived and the rootlocus method is applied to design compensator to improve the stability, response speed and accuracy of the system. Computer simulations are done to verify the goodness of the system design.

Normal operation when deeply submerged is a relatively easy task, and human operator control can often provide adequate performance. Near surface depthkeeping, on the other hand, is difficult to both man and machine. Because of the inherent limitation of the human operator, manual control may prove inadequate for near surface depthkeeping in some sea state. This paper describe the control algorithm of an automatic depth control system for submerged body that can be used for both near surface and deeply submerged depthkeeping operations. The computer simulations demonstrate the excellent depthkeeping performance of the controller under seaway effects.

This paper propose a optimal design method for robust notion controllers of underwater vehicles using the combined technique between classical and modern theories. The proposed method is presented which utilizes classical control methods to obtain a good robustness and modern control methods to set optimal gains. LQ, SVD, multivariable frequency analysis and BodeRoot Locus (BRL) plot are used.

Recently, the robot control systems are required to be more flexible and intelligent in order to execute more complex and delicate tasks. As an approach to such system, a PCbased robot control system is presented in this paper. The axis controllers are independently designed for each joint of robot manipulator and are supervised by a personal computer. Therefore, the almost system program including the control algorithm can be easily developed and modified in the PC's environment. This proposed control system was successfully applied to 4 axis SCARAtype robot and the associated S/W modules were developed. We can construct sensory robot systems by simply connecting the sensing devices to the PC and the study on this now being progressed.

본 논문에서는 다중로보트의 제어를 위하여 비젼 센서 등 여러 가지 센서신호를 처리할 수 있고 로봇의 충돌회피 및 협조제어를 할 수 있는 제어시스템을 개발하였다. 본 제어장치는 시스템 전체를 관리하며 언어 및 로보트 동작의 교시 그리고 자기진단 등의 기능을 하는 Supervisory Processor, vision에 대한 정보를 담당하는 로보트 제어 processor 등 여러개의 프로세서로 나누어 분산처리 구조를 갖도록 하여 확장성 및 유연성이 높은 시스템이 되도록 하였다. 실험적으로 본 시스템을 이용하여 로보트로 하여금 puzzle을 맞추는 작업을 수행시킴으로써 본 시스템의 우수성을 입증하였다.

The robot operation by teach playback is easy and was widely used for simple jobs performed by a simple robot manipulator. However, as robots and their control systems and tasks become more and more sophisticated, such a simple robot operation is no longer adequate and programming languages capable for the complicated systems and tasks are greatly needed. In this paper, a highlevel robotspecific programming language, SAITEL, is presented. It is an interpreter, based on Assembly, and has form similar to BASIC. SAITEL is easy to use for people who are not skilled programmers, and provides the capability to define robot task very conveniently. SAITEL was implemented on a direct drive SCARA robot developed in the Samsung Advanced Institute of Technology, and proved to be very useful for the operation of SCARAtype robots. It can be used also for other types of robots by slight modification.

오늘날, 생산성 증대와 품질 향상을 목적으로 산업용 로봇의 사용이 증대되면서 다양한 생산현장에서의 유연성과 적응성이 크게 요구되고 있는 실정이다. 그러나, 현재의 산업용 로봇은 endeffector의 dexterity의 부족으로 그 사용에 제한을 받고 있다. 이를 해결하기 위해 다지(multifingered), 다관절(articulated) 로봇손(robotic hands)의 도입이 필요하게 되었는데 이는 전체 로봇시스템의 능력과 유연성을 증가시킬 수 있을 것이다. 현재의 Flexible Manufacturing Workcell은 보통 한 대의 로봇팔과 많은 특수 용도의 고가의 tool로 구성되어 있으나 일반적인 모든 작업에 제한없이 모두 적용하기는 매우 어렵다. 그러므로, 다지, 다관절 로봇손의 개발 및 그 지능 제어의 실현은 작업에 따르는 불필요한 tool의 교환을 방지하여 비용과 시간을 절약함으로써 생산성을 극대화시킬 수 있다. 따라서, 본 연구에서는 Flexible Automation을 위한 다지, 다관절 로봇손의 설계 및 제작에 관한 방법을 제시하고자 한다.

This paper describes an experimental KnowledgeBased Control System, named KBCS, for manufacturing and assembly. The KBCS of five parts and function : database, knowledge acquisition, optimization, and graphic monitoring. The KBCS is utilized for a FMS which is of five machine centers and automatic assembly lines. Each machine can perform almost all manufacturing functions which some difference in efficiency. Buffers store temporarily the incoming components and the outing components. Parts arrive at assembly lines after many steps of manufacturing, and the transfer path and time are determined by procedural knowledge of control systems. Nine different incoming components are set up. The total control system is expected to perform four algorithms, timing algorithm ,sequencing algorithm, penalty algorithm, and cart algorithm. The construction of controller require basic components of manufacturing systems in which knowledges are formulated on the base of the results and the repeated simulation of KBCS with graphic monitoring system. Simulation results by KBCS are compared with those by the other rules of manufacturing.

This paper formulates a problem of analysis and design of serial production lines, closed with respect to the number of carriers available in the system for parts transportation between operations. For two machines  two buffers systems, the paper gives an asymptotic solution and shows that optimization of the system with respect to the number of carriers available and the capacity of the feedback buffer may lead to substantial improvements of system's performance.

This paper discusses the problem of designing an observation function as coarsest as possible in supervisory control of discrete event dynamic systems. Some algebraic properties of two sets consisting of observation functions for which the given desirable behavior is realizable are investigated:these sets with a partial ordering turn out not to possess largest elements. Two natural methods of obtaining an observation function, which is frequently coarser than the identity mapping, are also presented. An example illustrates the use of these methods.

This paper treats the computational time delay issue in designing digital control systems. The computational time delay margin, within which the closedloop stability is guaranteed, is analyzed using Rouche theorem. A PID control algorithm is proposed for compensating the computational time delay. Finally, the analyzed and the exact computational time delay margins are compared, and the performance of the proposed PID controller is shown through an illustrative example.

본 논문에서는 연속 및 이산 Lyapunov 방정식의 해의 고유치 및 트레이스의 범위를 시스템 행렬의 고유치 및 고유벡터 행렬을 이용하여 표시한다. 이산 시스템의 경우 시스템 행렬의 최대 특이치가 1보다 큰 경우나 연속 시스템의 경우 시스템 행렬의 대칭행렬이 불안정한 경우에도 상한 값이 항상 계산 가능한 범위가 제시된다. 본 논문에서 제시된 범위들은 몇가지 조건을 갖고 다른 문헌에서 제시된 것들 보다 정확하며, 더욱이 특정한 시스템 행렬에 대해서는 범위의 상한과 하한이 일치한다.

In this paper, we present a method of parameterizing LQG which simultaneously stabilizes a set of plant model by employing the well known separation principle and Hurwith polynomial. A conditions for existence of simultaneously stabilizing LQG for, given set of plant is drived. We show that suggested parameterization achieves the closed loop stability by fixed LQG for given set of plant.

In industrial multivariable plants, it is often the case that the plant outputs are detected not simultaneously but serially. In this paper, the problem of estimating the state vector of the plant based on the data obtained from such a detecting scheme is considered, and a special type of observer (referred to as a "serialsampling' type observer) which renews its internal states whenever a new data is obtained is proposed. It is proved that such an observer can be constructed for almost every sampling period if the plant is observable as a continuoustime multivariable system, and that the poles of the closedloop system using the serialsampling type observer consist of the poles of the observer and those of the state feedback system. The behaviors of the observer and the closedloop system are studied by simulation. The results of simulation indicate that a serialsampling type observer can estimate the state of the plant more accurately than the ordinary type observers and improve the closedloop performance.ance.

In the VSS control, reaching phase and chattering problem are the main disadvantages. A simple nonlinear sliding surface is proposed to cope with the reaching phase. The chattering problem is alleviated by almost continuous control with the compensation using the infinite closed loop.

A method of model reduction for reducing a higher order Ztransfer function to its lower order model is developed based on the Block  pulse function. The approach is following : I. Block  pulse function can be applied for Ztransfer function of linear digital system described by high order. II. To determined both the coefficients of the denominator and numerator of reduced model. The proposed method is simple for computation, can preserve the dynamic characteristic of the original model satisfactorily.

A new speed improvement method for quantitative superimposed EMG signal analysis to diagnose the neuromuscular dysfunction is described. The improvement is achieved through the use of efficient software and hardware signal processing techniques. The software approch is composed of the MANDF filter and HRWA algorithm which provides the optimal set and time delays ofselected templates. The hardware employs a TMS32OC25 DSP chip to execute the intensive calculation part. The purposed method is verified through a simulation with real templates which are obtained from needle EMG. As a results, the proposed method provides an overall speed improvement of 3240 times.

In this paper, the information capacity, the Integrate performance measure of digital radiography system has been quantitatively analyzed. The effect of various factors affecting the information capacity of the digital radiography system in multistage detection processes has been considered and analyzed In detail. The results show that the more Information capacity can be obtained with the smaller detector cell area, despite of the decreasing signal to noise ratio due to the reduced number of Xray photons per detector cell. And also has been Indicated that the Increase of the information capacity may be severely lost by the addition of electric noise in final readout stage.

Detection of left ventricular boundary for the functional analysis of LV(left ventricle)is obtained using automatic boundary detection algorithm based on dynamic programming method. This scheme reduces the edge searching time and ensures connective edge detection, since it does not require general edge operator, edge thresholding and linking process of other edge. detection methods. The left ventricular diastolic volume and systolic volume and systolic volume were computed after this automatic boundary detection, and these Volume data wm applied to analyze LV ejection fraction.

This paper describes an algorithm that can efficiently analyze the multichannel ECG signal using the frame. The input is a set of significant features (points) which have been extracted from an original sampled signal by using the splitandmerge algorithm. A signal from each channel can be hierarchical ADN/OR graph on the basis of the priori knowledge for ECG signal. The search mechanisms with some heuristics and the mixed paradigms of datadriven hypothesis formation are used as the major control mechanisms. The mutual relations among features are also considered by evaluating a score based on the relational spectrum. For recognition of morphologies corresponding to OR nodes, an hypothesis modification strategy is used. Other techniques such as instance, priority update of prototypes, and template matching facility are also used. This algorithm exactly recognized the primary points and supporting points from the multichannel ECG signals.

This paper presents a real time FECG signal monitoring system in which an adaptive multichannel noise canceller is implemented using a Texas Instruments TMS32020 digital signal processor. Abdominal ECG signal is applied as the desired output and 3 chest ECG signals as the reference input signals of the adaptive multichannel noise canceller whose coefficients are updated using the LMS algorithms.

In this paper we describe the shape measuring technique and system with a noncontractive sensor, composed of slitray projector and solidstate camera. For improving the accuracy and preventing measuring dead point, this sensor part is attached to the end of robot, and each sensing is executed after one step moving. By patching these sensing data, whole measuring data is constructed. The calibration between sensor and world coordinate is implemented through the specific calibration block by transformation matrix method. The result of experiment was satisfactory.

This paper describes a method of aisle following of a mobile robot using machine vision. As a navigation guide, Black strip painted on the lower part of wall of aisle is used. The offset of the vehicle position from the center of aisle and the heading angle are determined from the binary image of guide strip captured by a CCD camera. To remove the effect of noise. i.e. break of guide strip for the door or reflection of light, pixel sampling method together with consistency check of the incline for the sampled pixels is used.

When using robot manipulator to carry out autonomous tasks, the positioning accuracy of the robot manipulator relative to a reference coordinate frame is of greate importance. The task program, which is generated by offline CADsystem and used in actual robot positioning, may cause serious amount of the absolute positioning error of the robot manipulator. In this study, a robot performance test and calibration algorithms are proposed in order to improve the absolute positioning accuracy of the robot end effector. Experiments were also carried out by utilizing the HYUNDAI Robot AE 7601 and KIM2Tester, a three dimensional measurement system, which is developed in Robotics & Fluid Power Control Lab. at Korea Institute of Science and Technology.

Spent nuclear fuel is very dangerous substance emitting strong ionizing radiation which is harmful to human body. The remote handling of spent nuclear fuel is essential because people cannot access this substance without protecting radiation. To handle highly radioactive material or nuclear waste, many kinds of teleoperators such as master slave manipulator, electro mechanical manipulator, servo manipulator, mobile robot was developed. The distance and dimension of target object cannot be measured easily when highly radioactive material is handled by teleoperator because one should use lead glass or TV camera and monitor to protect radiation and see target object. During experiments on the remote handling of spent nuclear fuel by electro mechanical manipulator, we often felt that a distance and dimension measurement system is necessary to handle the objects which is in the highly radioactive environment, so we developed a system which is appropriate for this purpose.

The robot system which can handle a stream of randomly positioned parts on a conveyor belt system, is developed. It is composed of a PUMA 560 robot, a conveyor belt system and a vision system. The performance of the overall system is mainly dependent upon the robot and vision system interface technique. A vision algorithm is developed to determine the position, orientation and type of the part. Calibration procedure and the visiontorobot transformation are also proposed. Experimental results are then presented and discussed.

This paper presents a method of calculating the position and orientation of a polyhedron arbitrarily placed in 3D space using two cameras. We use key feature of the object and CAD data to solve the correspondence problem between two cameras' images.

In this study LLC and MAC program, which is the base of the communication software, is implemented. This program is tested in the MAP board for IBM PC. The LLC is of class 3, which provides the services of type 1 and 3. The MAC uses the Token Passing Bus Access Method of IEEE 802.4 standard. The LLC is implemented in C language, and MAC in C language and 80186 assembly language. MAC program takes advantage of MC 68824 which is the token bus controller produced by Motorola.

This paper describes the design of EFbus(ETRI Fieldbus), EFbus was designed from the interim results of international Fieldbus standardization in IEC (International Electrotechnical Commission) and proposals of many companies. The design was performed and described on the basis of three layered architecture. Application layer S/W runs on IBM PC and provides services which is similar ED MMS in MAP to user application Data linklayer runs on Intel's iDCX96 real time executive and uses centralized media accesscontrol method. Physical layer uses Manchester encoding & decoding, twisted pair fines and RS485 electrical standard.

In this paper, the architecture of a dataflow logic solving processor for programmable logic controller is proposed. As the proposed DFLSP (dataflow logic solving processor) is designed based on the dataflow architecture, it has inherently concurrent processing and data synchronization capabilities. The proposed DFLSP is adequate for high speed programmable logic controllers and gets rid of data synchronization problem in hardware level. The performance of the proposed DFLSP is analyzed using computer simulations and prototype hardware. With single processing element, the logic solving time is 144 usec per 1K steps of logic program and with eight processing elements, the logic solving time is 23 usec per 1K steps of logic program with reasonable assumptions.

본 EXOTHERMIC BATCH 공정은 두 개의 FLUID BED REACTOR로 구성되며 여기서 다종의 제품이 생산된다. 각 제품은 그 제품 자체의 특징 온도 패턴을 갖게 되며 허용한계를 벗어나지 않는 제품을 생산하기 위해서는 그 특징 온도 패턴대로 제어되어야 하며 실패할 경우 그 배치를 포기하여야만 한다. 초기의 배치제어시스템에서는 램프 제너레이터(RAMP GENERATOR) 기능이 있는 공기식 계기가 사용되었으며 제품을 변경하기 위해서는 각 제품마다 일일이 그 특징 패턴을 수동으로 입력시켜 주어야만 했다. 이 과정에서 한 개의 파라미터를 실수하면 사용할 수 없는 제품이 된다. 다종 제품의 화학공정에서는 품질과 생산량 향상에 의한 이득 증대에 그 목표를 두고 있으며, 이와 같은 목표를 달성하기 위한 해결책은 운전자의 개입을 감소시켜주는 제어시스템으로 개선하는 것이었다. 본 논문에서는 배치공정에 있어서의 프로그래머블 콘트롤러(PROGRAMMABLE CONTROLLER)의 응용에 관하여 논술하고자 한다.

Gunner primary sight(GPS) stabilization system lays line of sight(LOS) to find out a target and transmits informations to the fire control system (FCS). In a moving vehicle, accuracy of LOS and FCS depends on the design of GPS and servomechanism system. The heavy vibration of vehicle on the severe offroad environment degenerates the stabilization capability of GPS. In this study, to stabilize of elevation for GPS using the variable structure control, we derived the dynamic equation of GPS system and designed the variable structure controller. Computer simulation results fulfilled the static and dynamic stability of GPS using the variable structure control.

In this paper, an Inverted Pendelum is made and its state equation is formulated. Inverted Pendelum position and stabilization control is performed by variable structure system. Because conventional linear controller has difficulty in controlling inverted pendelum by precise method, VSS controller is chosen as a model to stabilize the Pendelum at the unstable equiblium point. In the experiment, the controller is implemented with the PC and it can stabilize the Inverted Pendelum.

Since linearized equations of notion have much modelling errors, robust controller for disturbances and noises Is necessary for autopilot. In this paper, notion equations for underwater vehicle with six degreeoffreedom are derived and linearized. And robust autopilot for this system is designed by using LQG/LTR methodology.

In the airfuel control of automotive engine to improve its efficiency, fuel economy and less emissions, conventional control methods using
$O_{2}$ sensor or the lean airfuel ratio sensor provide only open control in rich conditions. Control with a wide range airfuel sensor makes it possible to employ closed loop control for all engine conditions including rich combustion. With a wide range A/F sensor and A/F transfer functions, a PID control system is constructed which employs an learning scheme. A/F controller is designed which enables to improve the ability of its compensation for sensors and actuators, and its control operation is evaluated by computer simulation. 
This paper deals with the application of a Generalized Predictive Control (CPC) to a Pressurized Water Reactor (P.W.R) Nuclear Power Plant. Generalized Predictive Control is a sort of Explicit SelfTuning Control. Current selftuning algorithms lack robustness to prior choices of either deadtime (input time delay of a plant) or model order. GPC is shown by simulation studies to be superior to accepted selftuning techniques such as minimum variance and poleplacement from the viewpoint that it is robust to prior choices of deadtime or model order. In this paper a GPC controller is designed to control the P.W.R. nuclear power rlant with varying deadtime and through the designing procedure the designer is free from the constraint of knowing the exact deadtime. The controller is constructed based on the 2nd order linear model approximated in the vicinity of operating point. To ensure that this loworder model describes the complex real dynamics well enough for control purposes, model parameters are updated online with a Recursive Least Squares algorithm. Simulation results are successful and show the possibilities of the GPC control application to actual plants with varying or unknown deadtime.

A nonlinear feedback linearizing control method for a EMS(Electro Magnetic Suspension) system is proposed. After linearizing the system using the exact linearizing method, conventional linear system control theory has been applied. Computer simulations are carried out in order to compare the performance of the proposed controller with that of the existing controller designed by using Taylor series expansion around nominal points.

According to the recent increase of demands for multifunction and economics on hydraulic excavator, it is required that excavator should have simple operation, higher and operational efficiency. However, it is difficulty for current hydraulic system to satisfy demands fully. This study shows that new control system improves power transmission efficiency, work capability of engine and hydraulic system of current excavator.

The existing Swash Plate Type Axial Piston Motor rapidly drops efficiency at high speed in comparison with low speed. This is the reason why the motor was designed only in a viewpoint of minimum of power supply. But, in this paper, the motor which was optimally designed in a veiwpoint not only power supply but also load capacitancy, flow loss volume, axial stiffness and tilting stiffness keeps up high efficiency at high speed.

The purpose of this study is ID form the required force for measurements of the performances of the equipments or testpieces. For the generation of the required force, ft difference of pressures in each chamber of the hydraulic cylinder was controlled and Variable Structure Control theory was adopted to control it. Besides, observers Luenberger Observer and nonlinear Variable Structure Observer  were designed to estimate the derivative of the load pressure which is necessary ID determine the sliding surface in VSC theory. As a consequence of the computer simulation, it was shown that VSC had better performance than classical control theory(P, PD control) and VSO performed better than the Luenberger Observer at the load pressure control.

A position control system is developed for an electrohydraulic servo actuator with coaxial rotary spool, where the actuator is controlled by stepping motor. The position control system is utilized to develop the wireless remotely controlled crane system. And remote engine control system is also developed. Finally, to show the validity of this system, some experimental results and field test results am presented.

This paper describes an Electronicallycontrolled Power Steering system which is developed by the modification of a conventional power steering based on so called rotary valve technology. The steering effort is influenced by the electrohydraulic flow rate control of the pressurized oil to rotary valve. The vehicle speed and the steering angular velocity are used to calculate and output a signal to proportional flow rate control valve by the Electronic Control Unit. The improvement of the steering feel was satisfactory compared with that of the original conventional power steering.

This study presents the structure and the gate control algorithm of KAMOBOT (KAIST Mobile Robot). The mobile robot has a sixlegged, cylindrical configuration, each leg of which is equiped with a wheel at the bottom. The robot can go up and down stairs, go over obstacles, move along curvilinear paths and rotate around it's geometric center. Such maneuverability can be achieved by using only three electric motors.

This paper presents a path control method for mobile robot using neural network and a systematic method for the kinematic and dynamic modelling of a mobile robot. The robot finds its path deviation by taking the signals of an optical array sensor and determined its moving behaviors using neural net control method. A robot can be taught behaviors by changing the given patterns, in this work, Back Propagation rule is used as a learning method.

It is one of the essential task to determine the absolute location of mobile robot during its navigation. In this paper we propose an algorithm to calculate the distance and orientation of camera from landmark through the visual image of stripe typed landmark. Exact closed form solution of camera location is obtained with the correspondences from vertical line on mark plane to the intersection point of projected line with horizontal axis of image plane. It needs only one line image information, so that location determination can be processed in real time.

This paper proposes a new path planning method for obstacle avoidance of mobile robot. In order to achieve easy planning of the path, a simple representation of the empty space is achieved based on thinning algorithm. The proposed Planning technique facilitates the direct use of information obtained by camera. Comparing to the Vgraph method, the task of determining the shortest path from the resulting skeleton of empty space is optimized in terms of number of computation steps. The usefulness of the proposed method is ascertained by simulation.

In this paper, assuming that local range information is available, a collisionfree path planning algorithm for multiple mobile robots is presented by using Hopfield neural optimization network. The energy function of the network is built using the present position and the goal position of each robot as well as its local range information. The proposed algorithm has several advantages such as the effective passing around obstacles with the directional safety distance, the easy implementation of robot motion planning including its rotation, the realtime path planning capability from the totally localized computations of path for each robot, and the adaptivity on arbitrary environment since any special shape of obstacles is not assumed.

The mobile robot, named as KAEROT, is designed to go up and down stairs in nuclear facilities. To get a proper stable motion, kinematic modeling and analysis are seriously considered and new climbing algorithm is proposed focused on the stability. A couple of small wheels of one planetary wheel have to contact the surface ol stairs all the time to give the guarantee for stability and safety. To confirm the validity of the proposed algorithm, simulation is carried out. The results make evident of feasibility for the algorithm.

The communication structure in iron & steel plants are described. The specifications of the management network, the plant network and the field network are investigated. Data link methods between level 1, and level 2 are studied. Several network examples are illustrated.

This paper deals with the implementation of the IEEE 802.4 token passing bus protocol which is the world standard for LAN protocol. We are design station adaptor using MC68824 for data link layer, MC68194 for physical layer, and implemented by IEEE 802.4 and 802.2 type 3. We made three station adaptor and its performance is experimented.

MMS (Manufacturing Message Specification) is a communications standard of MAP (Manufacturing Automation Protocol), which is a Network Protocol for Industrial Automation. In the environment of MiniMAP, MMS is situated in the application layer and on the LLC (Logical Link Control) layer. In this paper, MMS software is implemented on the basis of ISO DIS 9506. The implemented software was tested with Token Passing Bus Network Interface Unit, which was designed in our laboratory and performs the actions of layer 1 and 2.

NC machine is xi expensive and valuable asset to a manufacturing enterprise. Consequently, a great deal of work has been done to maximize the productivity of this equipment. A major goal of all of these efforts is to achieve connection of this machine lo computers, integrating them into factory information system so that they can not only be driven, but cm be managed as well. This paper describes an approach to implement NC companion standard(NCCS) to MMS.

In this paper, the MiniMAP network interface units are implemented for the IBM PC, general purpose, PLC, and robot respectively. The token bus controller, MC68824 is used to realize the IEEE 802.4. A protocol for the message exchanges between some automated devices called MMS is implemented where 76 services of total 86 MMS services are included.

Manufacturing Message Specification(MMS) was sewed as an International Standard by International Standardization Organization(ISO). MMS is being accepted throughout the world as a solution to communications among multi vendor factory floor environments. This paper describes an implementation of MMS which operates on Application Layer of Open Systems Interconnection(OSI) 7 layer model. MMS was implemented on MSDOS in apersonal computer environment.

In this paper, a reliable control system structured with dual CPU modules and dual I/O modules is implemented as a means of achieving a highly reliable fault tolerant control system. For this, faults in the system modules are first examined, and a fault detection technique consisting of self diagnostic, comparison process, and exception processing is applied. Also reliability analysis is conducted for the discrete time Markov model with dual structure. It is shown quantitatively that the reliability is improved in the control system with dual structure in comparison with a system with single module structure.

Recently batch type cold rolling processes have been replaced by continuous annealing type processes for cold rolled sheets of mild steel and high strength steel in order to obtain higher productivity, labor saving. In the continuous annealing line, it is very important to maintain the target steel strip temperature at the exit side of each furnace. The automation system of continuous annealing line is based on a hierachical composition. This paper shows how to preset the set value of furnace temperature control for the heating section in a continuous annealing line. Saying in other words, this paper presents the development of an adaptive control approach to control the exit strip temperature in the continuous annealing line. There are three parts in this approach; one is a process modelling and another is recursive parameter estimation and the other is a design of temperature controller.

Computer simulation is carried out for passive, active, and semiactive suspension system. Each RMS and frequency response to road profile input is calculated for comparison and evaluation of the performance. The vibration analysis and active control of the quarter model of a vehicle suspension is studied in order to evaluate the alternative control laws. This paper derives an optimal closedloop feedback law for the semiactive suspension that justifies the clipped optimal approach.

Preprocessing technique of the measurement bearing data is presented to improve the target estimation accuracy for the bearingonly target notion analysis (TMA). Computer simulation is performed to compare with respect to the extended Kalman filter. By computer simulation, the target filter estimator with preprocessing Is both stable and robust to the measurement bearing noise.

To improve gun hit probability and to correct miss distance between target and bullet. It assumes that a radar tracks both the targets and bullets fired by a gun system. This papers describes an adaptive algorithm developed for processing the large number of radar measurements. The gunorder computation is enhanced by feedback from the ballistic estimator.

In the blade grinding process, Blade image is captured. It is captured in the environment with vibration, using monochrome CCD camera with high speed electronic shutter. The image is preprocessed using LoG filter and zero crossing. We used Hough transformation to detect straight lines from the preprocessed image. From the equations of detected lines, we calculated width between lines caused by grinders. This paper proposes automatic measurement of blade width to automate the process control of blade grinding line.

Optimal control systems for a radial magnetic bearing are proposed and tested. An electromagnet system model is developed to predict bearing forces as functions of changing coil currents and rotor position. Models of the multimass rotor supported by the magnetic bearings are developed and confirmed, and theoretical relationships are expressed in the form of state equation. As a result of computer simulation, the state feedback with optimal procedure is considered to be more effective for magnetic bearing systems than the classical method.

This paper describes the method on image acquisition and image processing in the turning process. The formation of discontinuous chips during highspeed oblique cutting without lubricant was observed by means of video camera recorder and stroboscope. The image processing technique for chip flow is described and the results are presented for variable feeds. It is concluded that experimental values of chip flow angle are similar to theoretical values of Stabler's rule.

This paper is prestage for getting EGI which can be used for modeling of an object. It discusses the construction of the vision processing system and its algorithm for getting needle diagram from tie object image. We realize the algorithm with monocular camera system, using Reflectance Map theory and photometric stereo method. We can calculate the surface normal at any point in the image if we take multiple images at the different lighting conditions. From the 3 images taken from different lighting conditions through the experiment, we get the needle diagrams of the sphere and the object. We confirm the validness of the surface, normal acquisition algorithm comparing the experimental needle diagram with the ideal one obtained from the surface normal of the known object.

This paper presents a method for Identification of a continuous time linear system parameters. We take the plant driven by percitently exciting input. To express the integral functions in terms of measured periodic output data. We use the Walsh function based on calsal functions. The linear algebraic equations for parameter identification is obtained. The present method Is simple and computationally advantageous.

This paper reports the design procedure and testing result of a servo driven pan/tilt device which is capable of tracking arbitrary movement of a specified target object. In order to achieve realtime acquisition of feedback signal, a 2 degreesoffreedom noncontact type displacement follower is used. The performance of the system is tested for different target velocities and control gains. The result of the research may provide an effective tool for visual transfer in the context of teleoperation.

Realtime implementations of active noise controller are proposed and tested. For compensation of time delay of feedback path, the nstep ahead prediction is applied. And predicting source noise and reflection noise respectively, reflection noise can be cancelled. For realtime processing, the DSP56001(Digital Signal Processor) is used. Experimental results show that the proposed controller is stable and of good performance.

Track While Scan(DS) system which can track the multitargets in dense target environment is designed. There are three tasks to be performed: I) Target Detection and 'plot' formation, ii) Plot to track association and, iii) Track updatement. The conventional approach has been to tackle each of these tasks separately. This paper outlines a method for jointly optimizing all the three tasks and presents implementation aspects.