• KIEE International Transaction on Systems and Control
• /
• v.12D no.1
• /
• pp.43-50
• /
• 2002

In the recent years neural networks have fulfilled the promise of providing model-free learning controllers for nonlinear systems; however, it is very difficult to guarantee the stability and robustness of neural network control systems. This paper proposes an adaptive neural network control for robot manipulators based on the radial basis function netwo.k (RBFN). The RBFN is a branch of the neural networks and is mathematically tractable. So we adopt the RBFN to approximate nonlinear robot dynamics. The RBFN generates control input signals based on the Lyapunov stability that is often used in the conventional control schemes. The saturation function is also chosen as an auxiliary controller to guarantee the stability and robustness of the control system under the external disturbances and modeling uncertainties.

• Phonetics and Speech Sciences
• /
• v.11 no.2
• /
• pp.57-64
• /
• 2019

In this paper, we construct a Korean text-to-speech system using the Merlin toolkit which is an open source system for speech synthesis. In the text-to-speech system, the HMM-based statistical parametric speech synthesis method is widely used, but it is known that the quality of synthesized speech is degraded due to limitations of the acoustic modeling scheme that includes context factors. In this paper, we propose an acoustic modeling architecture that uses deep neural network technique, which shows excellent performance in various fields. Fully connected deep feedforward neural network (DNN), recurrent neural network (RNN), gated recurrent unit (GRU), long short-term memory (LSTM), bidirectional LSTM (BLSTM) are included in the architecture. Experimental results have shown that the performance is improved by including sequence modeling in the architecture, and the architecture with LSTM or BLSTM shows the best performance. It has been also found that inclusion of delta and delta-delta components in the acoustic feature parameters is advantageous for performance improvement.

• Proceedings of the Korea Technical Association of the Pulp and Paper Industry Conference
• /
• 2004.04a
• /
• pp.86-96
• /
• 2004

Identification of dynamics of short circulation during grade change operations in paper mills is very important for the effective plant operation. In the present study a prediction method of One Pass Retention(OPR) is proposed based on the neural network. The present method is used to analyze the dynamics of short circulation during grade change. Properties of the product paper largely depend upon the change in the OPR. In the present study the OPR is predicted from the training of the network by using grade change operation data. The results of the prediction are applied to the modeling equation to give flow rates and consistencies of short circulation.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1995.04b
• /
• pp.198-203
• /
• 1995

In the paper, we have proposed a new technique to detemine the initial billet for the forged products using a function approximation in neural network. A three-layer neural network is used and a back propagation algorithm is employed totrain the network. An optimal billet which satisfied the forming limitation, minimum of incomplete filling in the die cavity, load and energyas well as more uniform distribution of effective strain, is determined by applying the ability of function approximation of te neural network. The amount of incomplete filling in the die, load and forming energyas well as effective strain are measured by the rigid-plastic finite element method. The new technique is applied tofind the optimal billet size for the axisymmetric rib-web product in hot forging. This would reduce the number of finite element simulation for determing the optimal billet of forging products, further it is usefully adapted to physical modeling for the forging design.

• Proceedings of the KIEE Conference
• /
• 2004.11c
• /
• pp.261-264
• /
• 2004

In this paper, we identify the memory effect of high power(125W) laterally diffused metal oxide-semiconductor(LDMOS) RF Power Amplifier(PA) by two tone IMD measurement. We measure two tone IMD by changing the tone spacing and the power level. Different asymmetric IMD is founded at different center frequency measurements. We propose the Tapped Delay Line-Neural Network(TDNN) technique as the modeling method of LDMOS PA based on two tone IMD data. TDNN's modeling accuracy is highly reasonable compared to the memoryless adaptive modeling method.

• The Transactions of the Korean Institute of Electrical Engineers P
• /
• v.62 no.2
• /
• pp.65-71
• /
• 2013

Nonlinear loads with harmonics exist in an actual power system where harmonic currents make voltage distortion. The sum of reactive power measured at individual load is different from the measured reactive power at a bus in a power system with linear and non-linear loads. In this study, ANN(artificial neural network) load modeling technique with consideration of harmonics is introduced for more accurate component load modeling and an impact coefficient is proposed for aggregation of component loads. Results of this research show more accurate load modeling method. Since precise data for power system analysis can be acquired, the proposed method will be used for power system planning and maintenance.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.16 no.7
• /
• pp.1322-1331
• /
• 1992

The conventional neural networks are built without considering the underlying structure of the problems. Hence, they usually contain redundant weights and require excessive training time. A novel neural network structure is proposed for symmetric problems, which alleviate some of the aforementioned drawback of the conventional neural networks. This concept is expanded to that of the constrained neural network which may be applied to general structured problems. Because these neural networks can not be trained by the conventional training algorithm, which destroys the weight structure of the neural networks, a proper training algorithm is suggested. The illustrative examples are shown to demonstrate the applicability of the proposed idea.

• 제어로봇시스템학회:학술대회논문집
• /
• 1991.10a
• /
• pp.1017-1021
• /
• 1991

This paper is to explore the potential use of the modeling capacity of neural networks for control applications. The tasks are carried out by two neural networks which act as a plant identifier and a system controller, respectively. Using information stored in the identification network control action has been developed. Without supervising control signals are generated by a gradient type iterative algorithm.

• 제어로봇시스템학회:학술대회논문집
• /
• 1996.10b
• /
• pp.139-142
• /
• 1996

This paper presents a method of controlling contact force for deburring tasks. The cope with the nonlinearities and time-varying properties of the robot and the environment, a neural network control theory is applied to design the contact force control system. We show that the contact force between the hand and the contacting surface can be controlled by adjusting the command velocity of a robot hand, which is accomplished by the modeling of a robot and the environment as Mass-Spring-Damper system. Simulation results are shown.

• 제어로봇시스템학회:학술대회논문집
• /
• 1992.10a
• /
• pp.218-223
• /
• 1992

This paper describes a learning method of neural network controllers. The learning method improves the performance of indirect learning mechanism in the neuro-control of nonlinear systems. To precisely identify dynamic characteristics of the plant by utilizing a limited prior information we propose a new energy function which takes advantage of the proportional relationship between outputs of the plant and those of neural networks.