• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.45 no.4
• /
• pp.574-581
• /
• 1996

The neural network approach has been shown to be a general scheme for nonlinear dynamical system identification. Unfortunately the error surface of a Multilayer Neural Networks(MNN) that widely used is often highly complex. This is a disadvantage and potential traps may exist in the identification procedure. The objective of this paper is to identify a nonlinear dynamical systems based on Self-Organized Distributed Networks (SODN). The learning with the SODN is fast and precise. Such properties are caused from the local learning mechanism. Each local network learns only data in a subregion. This paper also discusses neural network as identifier of nonlinear dynamical systems. The structure of nonlinear system identification employs series-parallel model. The identification procedure is based on a discrete-time formulation. Through extensive simulation, SODN is shown to be effective for identification of nonlinear dynamical systems. (author). 13 refs., 7 figs., 2 tabs.

• Proceedings of the KIEE Conference
• /
• 1999.07g
• /
• pp.2932-2934
• /
• 1999

In this paper, a scheme for recognition of handwritten digits using a multilayer neural network trained with the back-propagation algorithm using generalized delta rule is proposed. The neural network is trained with hand written digit data of different writers and different styles. One of the purpose of the work with neural networks is the minimization of the mean square error(MSE) between actual output and desired one. The back-propagation algorithm is an efficient and very classical method. The back-propagation algorithm for training the weights in a multilayer net uses the steepest descent minimization procedure and the sigmoid threshold function. As an error rate is reduced, recognition rate is improved. Therefore we propose a method that is reduced an error rate.

• Proceedings of the KIEE Conference
• /
• 1996.07b
• /
• pp.808-810
• /
• 1996

In this paper, neural network, which has learning capability, is used for fault type classification and fault section estimation for high speed relaying. The potential of the neural network approach is demonstrated by simulation using ATP. The instantaneous values of voltages and currents are used the inputs of neural networks. This approach determines the fault section directly. In this paper, back-propagation network(BPN) is used for fault type classification and fault section estimation and can use for high speed relaying because it determines fault section within a few msec.

• Proceedings of the KIEE Conference
• /
• 1996.07b
• /
• pp.811-813
• /
• 1996

High impedance fault can not be easily detected by conventional method. But if it would not be detected and cleared quickly, it can result in fires, and electric shock. In this paper, neural network, which has learning capability, is used for high impedance fault detector. The potential of the neural network approach is demonstrated by simulation using KEPCO's measured data. The instantaneous values and frequency spectrum of current are respectively used as the inputs of neural networks. Also, the methods using combined data to exploit the advantage of each data are proposed. In this paper, back-propagation network(BPN) is used for high impedance fault detector and can use for high speed relay because it detects faults within 1 cycle.

• Proceedings of the KIEE Conference
• /
• 1993.07a
• /
• pp.248-250
• /
• 1993

In this paper, we assume that the dynamics of DC motor and nonlinear load are unknown. We train the inverse dynamic model of DC motor and nonlinear load using the neural network and construct speed control system based on the traind dynamic model and current control mode. Speed prediction scheme using neural network is also proposed the alleviate the time delay effect caused by the computation time of neural network. Simulation results show good performances of the control system. Finally, hardware configuration of the control system is outlined.

• Proceedings of the KIEE Conference
• /
• 1996.07b
• /
• pp.939-942
• /
• 1996

This paper presents a stable learning algorithm for diagonal recurrent neural network(DRNN). DRNN is applied to a problem of controlling nonlinear dynamical systems. A architecture of DRNN is a modified model of the Recurrent Neural Network(RNN) with one hidden layer, and the hidden layer is comprised of self-recurrent neurons. DRNN has considerably fewer weights than RNN. Since there is no interlinks amongs in the hidden layer. DRNN is dynamic mapping and is better suited for dynamical systems than static forward neural network. To guarantee convergence and for faster learning, an adaptive learning rate is developed by using Lyapunov function. The ability and effectiveness of identifying and controlling a nonlinear dynamic system using the proposed algorithm is demonstrated by computer simulation.

• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.132-134
• /
• 2003

The likelihood ratio, logistic regression and artificial neural networks methods are applied and verified for analysis of landslide susceptibility in Yongin, Korea using GIS. From a spatial database containing such data as landslide location, topography, soil, forest, geology and land use, the 14 landsliderelated factors were calculated or extracted. Using these factors, landslide susceptibility indexes were calculated by likelihood ratio, logistic regression and artificial neural network methods. Before the calculation, the study area was divided into two sides (west and east) of equal area, for verification of the methods. Thus, the west side was used to assess the landslide susceptibility, and the east side was used to verify the derived susceptibility. The results of the landslide susceptibility analysis were verified using success and prediction rates. The v erification results showed satisfactory agreement between the susceptibility map and the exis ting data on landslide locations.

• Proceedings of the KIEE Conference
• /
• 1995.07a
• /
• pp.216-218
• /
• 1995

In this article a neural network adaptive observer is proposed and applied to the case of induction motor control. The high performance vector control drives require exact knowledge of rotor flux. Because rotor time constant is needed to observe rotor flux, the accurate estimation of rotor time constant is important. For these problems, proposed observer which comprises neural network flux observer and neural network torque observer is trained to learn the flux dynamics and torque dynamics and subject to further on-line training by means of a backpropagation algorithem. Therefore it has been shown that the robust control of induction motor neglects the rotor time constant variations.

• Proceedings of the KWS Conference
• /
• 1992.10a
• /
• pp.91-95
• /
• 1992

The resistance spot welding process has been extensively used for joining of sheet metals, which are subject to variation of many process variables. Many qualitive analyses of sampled process variables have been attempted to predict nugget size. In this paper, dynamic resistance and electrode movement signal which is a good indicative of the nugget size was examined by introducing an artificial neural network estimator. An artificial neural feedforward network with back-propagation of error was applied for the estimation of the nugget size. The prediction by the neural network is in good agreement with the actual nugget size for resistance spot welding of galvanized steel. The results are quite promising in that the quantitative estimation of the invisible nugget size can be achieved without conventional destructive testing of welds.

• Journal of the Korean Society for Precision Engineering
• /
• v.21 no.5
• /
• pp.55-63
• /
• 2004

This paper presents neural network based controller using the feedback error loaming technique for a heavy load pointing system. Also the mathematical model was developed to analyze heavy load pointing system. The control scheme consists of a feedforward neural network controller and a fixed-gain feedback controller. This neural network controller is trained so as to make the output of the feedback controller zero. The proposed controller is compared with the conventional PI controller through simulations, and the results show that the pointing accuracy of the proposed control system are improved against the disturbance induced by vehicle running on the bump course.