• 제어로봇시스템학회:학술대회논문집
• /
• 1992.10a
• /
• pp.224-229
• /
• 1992

We investigate a neural network as a dynamic system controller when system characteristics are abruptly changing. The shape of sigmoid functions are determined by autotuing method for the optimum sigmoid function of the neural networks. By using information stored in the identifying network a novel algorithm that can adapt the control action of the controller has been developed. Robustness can be seen from its ability to adjust large variations of parameters. The potential of the proposed method is demonstrated by simulations.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.179-179
• /
• 2000

An adaptive neural network compensator for stick-slip friction phenomena in servo systems is proposed to supplement the traditionally available position and velocity control loops for precise motion control. The neural network compensator plays a role of canceling the effect of nonlinear slipping friction force. This enables the mechatronic systems more precise control and realistic design in the digital computer. It was confirmed that the control accuracy is more improved near zero velocity and the points of changing the moving direction through numerical simulation

• Transactions of the Korean Society of Machine Tool Engineers
• /
• v.13 no.3
• /
• pp.16-23
• /
• 2004

Thermally induced errors of machine tools have been recognized as one of the most important issues in precision machining. This is probably the most formidable obstacle to obtain high level of machining accuracy. To this regard, the experimental compensation methodologies such as software-based method or origin shift of machine tool axes have been suggested. In this research, to cope with thermal deformation, a model based correction was carried out with the function of an external machine coordinate shift. Models with multi-linear regression or neural network were investigated to selected a good one for thermal compensation. Consequently, multi-linear regression model combined with origin shift was verified good enough form the machining of dot matrices of plate with ball end milling.

• Journal of Power Electronics
• /
• v.9 no.3
• /
• pp.403-409
• /
• 2009

Automatically tuned passive filters can improve power quality to a great extent in power systems. A novel three-phase shunt auto-tuned filter is designed to effectively compensate source current harmonics and to provide reactive power required by the non-linear load, which draws a highly reactive, harmonic-rich current from the supply. An artificial neural network (ANN) based controller selects filter component values in accordance with reactive power requirement and harmonic compensation. Traditional passive filters are permanently connected to the system and draw large amounts of source current even under light load conditions. By using auto-tuned filters, the passive filter components can be controlled according to load variations and, hence, draw only required source currents. The selection is done by the ANN with the help of a properly tuned knowledge base to provide instantaneous compensation using a digital controller.

• Journal of Institute of Control, Robotics and Systems
• /
• v.2 no.4
• /
• pp.312-317
• /
• 1996

This paper proposes a neural network controller with the computed torque method. The neural network is used not to learn the inverse dynamic model but to compensate the uncertainties of robotic manipulators. When training the neural network, we use the signals present in the proposed controller, which is simpler than that proposed by Ishiguro et al., whose teaching signals of the neural network come from the robot model.

• Journal of Institute of Control, Robotics and Systems
• /
• v.3 no.3
• /
• pp.252-258
• /
• 1997

This paper presents neural networks for hybrid position/force control which is a type of position and force control for robot manipulators. The performance of conventional hybrid position/force control is excellent in the case of the exactly-known dynamic model of the robot, but degrades seriously as the uncertainty of the model increases. Hence, the neural network control scheme is presented here to overcome such shortcoming. The introduced neural term is designed to learn the uncertainty of the robot, and to control the robot through uncertainty compensation. Further more, the learning rule of the neural network is derived and is shown to be effective in the sense that it requires neither desired output of the network nor error back propagation through the plant. The proposed scheme is verified through the simulation of hybrid position/force control of a 6-dof robot manipulator.

• Proceedings of the Korea Society for Simulation Conference
• /
• 2001.10a
• /
• pp.116-122
• /
• 2001

A novel approach based on WaveARX neural network observer is proposed far the fault detect of a class of nonlinear systems which consist of known linear part and unknown nonlinear part. A linear observer is first designed, then a nonlinear compensation term in the nonlinear observer is estimated by using a deconvolution method. The WaveARX network is used to model the obtained compensation term. At last, the residual fur fault detection is generated based on the analysis of the upper bound approximate error. Simulation results have shown the feasibility and effectiveness of the method.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.114.5-114
• /
• 2001

In this paper, the model for personal navigation system using low-cost inertial sensors and error compensation method with GPS are proposed. Simulation is accomplished for the performance test. WNS(Walking Navigation System) is a kind of personal navigation using the number of a walk, stride and azimuth. Because the accuracy of these variables determines the navigational performance, computational methods have been investigated. The step is detected using the motion pattern by walking motion, stride is determined by neural network and azimuth is calculated with gyro´s output. The neural network filters off unnecessary motions. However, error compensation method is needed, because the error of navigation information increases with time ...

• International Journal of Precision Engineering and Manufacturing
• /
• v.6 no.1
• /
• pp.31-35
• /
• 2005

Higher productivity requires high-speed motion of machine tool axes. The iron core linear DC motor (LDM) is widely accepted as a viable candidate for high-speed machine tool feed unit. LDM, however, has two inherent disturbance force components, namely cogging and thrust force ripple. These disturbance forces directly affect the tracking accuracy of the feeding system and must be eliminated or reduced. In order to reduce motor ripple, this research adapted the feedforward compensation method and neural network control. Experiments carried out with the linear motor test setup show that these control methods are effective in reducing motor ripple.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2004.10a
• /
• pp.358-362
• /
• 2004

Higher productivity requires high-speed motion of machine tool axes. The iron core linear DC motor (LDM) is widely accepted as a viable candidate for high-speed machine tool feed unit. LDM, however, has two inherent disturbance force components, namely cogging and thrust force ripple. These disturbance forces directly affect the tracking accuracy of the feeding system and must be eliminated or reduced. In order to reduce motor ripple, this research adapted the feedforward compensation method and neural network control. Experiments carried out with the linear motor test setup show that these control methods are effective in reducing motor ripple.