• 제어로봇시스템학회:학술대회논문집
• /
• 1992.10a
• /
• pp.692-697
• /
• 1992

This study concerns about the orbit prediction and orbit determination of Korean future communication satellite, called "Moogunghwa", which will be motioned in the geo-stationary orbit. Perturbation effect on the satellite orbit due to nonspherical geopotential term, lunar and solar gravity, drag force of the atmosphere and solar radiation pressure was investigated. Cowell's method is used for orbit prediction. Orbit determination was performed by using EKF which is suitable for real-time orbit determination. The result shows that the characteristics of the satellite orbit has drift. So the periodic control time and control value in the view of the periodic of error can be provided. The orbit determination demonstrated the effectiveness since the convergence performance on the position and velocity error , and state error standard deviation is reasonable.easonable.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2000.04a
• /
• pp.171-175
• /
• 2000

Neural networks are in the framework of sensorbased tracking control of robot manipulators. They learn by practice movements the relationship between PSD (an analog Position Sensitive Detector) sensor readings for target positions and the joint commands to reach them. Using this configuration, the system can track or follow a moving or stationary object in real time. Furthermore, an efficient neural network architecture has been developed for real time learning. This network uses multiple sets of simple backpropagation networks one of which is selected according to which division (corresponding to a cluster of the self-organizing feature map) in data space the current input data belongs to. This lends itself to a very training and processing implementation required for real time control.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2000.10a
• /
• pp.108-112
• /
• 2000

Neural networks are used in the framework of sensorbased tracking control of robot manipulators. They learn by practice movements the relationship between PSD (an analog Position Sensitive Detector) sensor readings for target positions and the joint commands to reach them. Using this configuration, the system can track or follow a moving or stationary object in real time. Furthermore, an efficient neural network architecture has been developed for real time learning. This network uses multiple sets of simple backpropagation networks one of which is selected according to which division (corresponding to a cluster of the self-organizing feature map) in data space the current input data belongs to. This lends itself to a very training and processing implementation required for real time control.

• Proceedings of the KIEE Conference
• /
• summer
• /
• pp.1343-1345
• /
• 2004

The inherent pulsating force generation makes the control of switched-reluctance motor quite complicate. So in this paper, an orthogonal scheme that can be applied to the control of LSRM will be presented. By using this scheme, the motor reluctance forces, which are magnetically decoupled and position dependent, can be projected onto sets of stationary axes that are aligned with the motor fixed secondary poles. Hence the electromagnetic forces can be controlled not only for driving the LSRM at its propulsive direction but also for alleviating the load at its normal direction. Simulation results will be provided to show the validity of the proposed scheme.

• Journal of Institute of Control, Robotics and Systems
• /
• v.14 no.7
• /
• pp.702-710
• /
• 2008

In this paper, the observability of the INS/GPS integrated system for a land vehicle is analyzed on measurements and different filters with respect to the measurements are designed. In the stationary case, it is shown that horizontal accelerometer biases and vertical attitude errors and gyro biases are unobservable. An 8-state filter is designed based on the observability analysis. When GPS signal is available, a 15-state filter is used with position and velocity measurements. To estimate the INS errors even in the case that GPS signal is blocked a filter is designed in consideration of the non-holonomic constraints of a land vehicle. In this case, the horizontal position and velocity errors and vertical attitude error are unobservable. However, a 12-state filter including the velocity states is designed to estimate the accelerometer biases. When GPS signal recovers, a 9-state filter is used excluding the sensor biases. This paper presents a multi-dimensional filter that switches the four filters according to the usable measurements and maneuver environments. A simulation is carried out to verify the performance of the proposed filter.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.10 no.1
• /
• pp.87-94
• /
• 2005

This paper is about the compensation of phase angle error and amplitude difference in stationary 2 pole coordinates by the installation variation of sensor at low cost high resolution digital hybrid encoder. To solve the problem of amplitude difference at the existing hybrid encoder, we normalized the relative magnitude of the two analog signals. and also to solve the problem of installation variation when installed the sensor, we rotate the stationary 2 pole coordinates to compensate the location error of sensor. and we used and tested the QEP function and A/D port in DSP(TMS320F2812) to verify the mentioned proposed method.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1997.11a
• /
• pp.259-262
• /
• 1997

This paper presents a visual servoing of RV-M2 robot manipulators to track and grasp moving object, using pruned dynamic recurrent neural networks(DRNN). The object is stationary in the robot work space and the robot is tracking and grasping the object by using CCD camera mounted on the end-effector. In order to optimize the structure of DRNN, we decide the node whether delete or add, by mutation probability, first in case of delete node, the node which have minimum sum of input weight is actually deleted, and then in case of add node, the weight is connected according to the number of case which added node can reach the other nodes. Using evolutionary programming(EP) that search the struture and weight of the DRNN, and evolution strategies(ES) which train the weight of neuron, we pruned the net structure of DRNN. We applied the DRNN to the Visual Servoing of a robot manipulators to control position and orientation of end-effector, and the validity and effectiveness of the pro osed control scheme will be verified by computer simulations.

• Journal of the Society of Naval Architects of Korea
• /
• v.42 no.2 s.140
• /
• pp.98-104
• /
• 2005

The mathematical models of hydrodynamic force and moment acting on a ship at low speed range should be established differently from the ones at nominal cruising speed range since a ship moves with large drift angle or rotates in a stationary position. We modified widely used Yoshimura's cross flow model in order to apply the system identification method to estimate parameters in the model. The apparatus and the procedure of free running model test were suggested so that the parameters in the model be estimated. The validity of our proposing modified model and test procedure was confirmed by comparison with the results of simulated model test.

• Journal of Korean Institute of Industrial Engineers
• /
• v.31 no.2
• /
• pp.106-119
• /
• 2005

Uncertainties inherent in customer demand patterns make it difficult for supply chains to achieve just-in-time inventory replenishment, resulting in loosing sales opportunity or keeping excessive chain wide inventories. In this paper, we propose two intelligent adaptive inventory control models for a supply chain consisting of one supplier and multiple retailers, with the assumption of information sharing. The inventory control parameters of the supplier and retailers are order placement time to an outside source and reorder points in terms of inventory position, respectively. Unlike most extant inventory control approaches, modeling the uncertainty of customer demand as a stationary statistical distribution is not necessary in these models. Instead, using a reinforcement learning technique, the control parameters are designed to adaptively change as customer demand patterns change. A simulation based experiment was performed to compare the performance of the inventory control models.

• Journal of Korean Institute of Industrial Engineers
• /
• v.22 no.2
• /
• pp.165-177
• /
• 1996

This paper presents a navigation method for a mobile robot which can follow a path drawn on a monitor screen. The current position and heading direction of the mobile robot are obtained using a single camera from a single image of the guide mark projected on a convex mirror. This proposed method eliminates the need to rotate or tilt the camera to track the guide mark, because the convex mirror and a stationary camera provide a panoramic view of the surrounding area. As guide marks, fluorescent lamps on the ceiling, door frames, or any other natural line segment can be used without adding any artificial elements to them.