• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.7 no.1
• /
• pp.99-118
• /
• 2013

Electrocardiograms (ECGs) are widely used by clinicians to identify the functional status of the heart. Thus, there is considerable interest in automated systems for real-time monitoring of arrhythmia. However, intra- and inter-patient variability as well as the computational limits of real-time monitoring poses significant challenges for practical implementations. The former requires that the classification model be adjusted continuously, and the latter requires a reduction in the number and types of ECG features, and thus, the computational burden, necessary to classify different arrhythmias. We propose the use of adaptive learning to automatically train the classifier on up-to-date ECG data, and employ adaptive feature selection to define unique feature subsets pertinent to different types of arrhythmia. Experimental results show that this hybrid technique outperforms conventional approaches and is therefore a promising new intelligent diagnostic tool.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.62 no.11
• /
• pp.1621-1627
• /
• 2013

This paper compares power losses between voltage controlled before and after using power conversion device in AC feeding system. For this purpose we present voltage control procedures and criteria and model high speed line and train using PSCAD/EMTDC to compare power losses in various feeding condition. Power losses of the simulation result in power control before and after in single point feeding system was reduced maximum 0.37 MW(23.8 %) and average 0.23 MW(20.5 %) when one vehicle load operates maximum load condition. When three vehicles operate maximum load condition in one feeder section, power losses after voltage control was reduced 1.03 MW(49.5%) compared to before voltage control. And, power loss of parallel feeding system is reduced the average 0.08 MW(7.2 %) compared to the single feeding system. In conclusion, adaptive voltage control method using power conversion device can reduce power losses compared with existing method.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.1115-1120
• /
• 2003

The present paper shows the possible development of particle swarm optimization (PSO) based fuzzy-neural networks (FNN) which can be employed as an important building block in real life robot systems, controlled by voice-based commands. The PSO is employed to train the FNNs which can accurately output the crisp control signals for the robot systems, based on fuzzy linguistic spoken language commands, issued by an user. The FNN is also trained to capture the user spoken directive in the context of the present performance of the robot system. Hidden Markov Model (HMM) based automatic speech recognizers are developed, as part of the entire system, so that the system can identify important user directives from the running utterances. The system is successfully employed in a real life situation for motion control of a redundant manipulator.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.640-645
• /
• 2004

This paper proposes a direct adaptive control method using self-recurrent wavelet neural network (SRWNN) for stable path tracking of mobile robots. The architecture of the SRWNN is a modified model of the wavelet neural network (WNN). Unlike the WNN, since a mother wavelet layer of the SRWNN is composed of self-feedback neurons, the SRWNN has the ability to store the past information of the wavelet. For this ability of the SRWNN, the SRWNN is used as a controller with simpler structure than the WNN in our on-line control process. The gradient-descent method with adaptive learning rates (ALR) is applied to train the parameters of the SRWNN. The ALR are derived from discrete Lyapunov stability theorem, which are used to guarantee the stable path tracking of mobile robots. Finally, through computer simulations, we demonstrate the effectiveness and stability of the proposed controller.

• Journal of Institute of Control, Robotics and Systems
• /
• v.21 no.6
• /
• pp.518-523
• /
• 2015

This paper proposes a gait phase classifier using a Recurrent Neural Network (RNN). Walking is a type of dynamic system, and as such it seems that the classifier made by using a general feed forward neural network structure is not appropriate. It is known that an RNN is suitable to model a dynamic system. Because the proposed RNN is simple, we use a back propagation algorithm to train the weights of the network. The input data of the RNN is the lower body's joint angles and angular velocities which are acquired by using the lower limb exoskeleton robot, ROBIN-H1. The classifier categorizes a gait cycle as two phases, swing and stance. In the experiment for performance verification, we compared the proposed method and general feed forward neural network based method and showed that the proposed method is superior.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 2008.04a
• /
• pp.114-119
• /
• 2008

A refined three-dimensional finite element interaction model between the high-speed train and railway bride deck has been developed in the present study. Analytical predictions of vertical deflections for a railway bridge are compared with in-situ test results and a good agreement is achieved. Then, input variables employed in the analytical comparisons are selected as random variables for the limit state functions. followed by risk assessment. For this purpose, a linear adaptive weighted response surface method has been developed and applied. A typical railway bridge has been selected and the limit state functions are employed from UIC and Korean specifications in the comparative studies. The results reveal that Korean specifications give significantly risky reliability indices in comparison with UIC specifications. It is thus encouraged from the above that the present linear adaptive weighted response surface method can be an alternative for the fast estimation of nonlinear structural systems.

• International Journal of Naval Architecture and Ocean Engineering
• /
• v.10 no.3
• /
• pp.235-249
• /
• 2018

The Artificial Neural Network (ANN) model has been known as one of the most effective theories for automatic ship berthing, as it has learning ability and mimics the actions of the human brain when performing the stages of ship berthing. However, existing ANN controllers can only bring a ship into a berth in a certain port, where the inputs of the ANN are the same as those of the teaching data. This means that those ANN controllers must be retrained when the ship arrives to a new port, which is time-consuming and costly. In this research, by using the head-up coordinate system, which includes the relative bearing and distance from the ship to the berth, a novel ANN controller is proposed to automatically control the ship into the berth in different ports without retraining the ANN structure. Numerical simulations were performed to verify the effectiveness of the proposed controller. First, teaching data were created in the original port to train the neural network; then, the controller was tested for automatic berthing in other ports, where the initial conditions of the inputs in the head-up coordinate system were similar to those of the teaching data in the original port. The results showed that the proposed controller has good performance for ship berthing in ports.

• Journal of Korea Multimedia Society
• /
• v.21 no.2
• /
• pp.98-107
• /
• 2018

In this paper, we propose a method to classify the skeletal maturity with a small amount of hand wrist X-ray image using deep learning-based meta-learning. General deep-learning techniques require large amounts of data, but in many cases, these data sets are not available for practical application. Lack of learning data is usually solved through transfer learning using pre-trained models with large data sets. However, transfer learning performance may be degraded due to over fitting for unknown new task with small data, which results in poor generalization capability. In addition, medical images require high cost resources such as a professional manpower and mcuh time to obtain labeled data. Therefore, in this paper, we use meta-learning that can classify using only a small amount of new data by pre-trained models trained with various learning tasks. First, we train the meta-model by using a separate data set composed of various learning tasks. The network learns to classify the bone maturity using the bone maturity data composed of the radiographs of the wrist. Then, we compare the results of the classification using the conventional learning algorithm with the results of the meta learning by the same number of learning data sets.

• Convergence Security Journal
• /
• v.14 no.6_1
• /
• pp.113-120
• /
• 2014

Recent changes made lack of volunteers is ready for a lot of the black students lack basic physical fitness and to increase this by itself is a lot to announce the final selection from the first notes, the secondary Fitness NCO selection process and interview process. The trend of course, but also to life as a career soldier passed the sergeant exam showed the important need is ready to run out of the old basic fundamentals that can be said that a lot of beginners executives. Therefore, we propose an appropriate national security with the mental and physical knowledge that our military needs the basic data ingest through analysis of model train to make excellent NCO personnel by fitness programs.

• The KSFM Journal of Fluid Machinery
• /
• v.16 no.3
• /
• pp.32-38
• /
• 2013

This paper deals with the study on the rotordynamic and experimental analysis of turbine-generator system connected with a magnetic coupling. Although magnetic coupling has been used to torque transmission of chemical processing pump rotating at under 3,600rpm, magnetic coupling in this study is applied to high-speed turbine-generator system using a working fluid that is refrigerant such as ammonia or R-124a. Results of rotordynamic design analysis are as follows. The first, shaft diameter nearest to outer hub of magnetic coupling has a big effect on the $1^{st}$ critical speed of generator rotor. The second, if the $1^{st}$ critical speeds of turbine rotor and generator rotor have enough to separation margin in comparison to rated speed, the $1^{st}$ critical speed of turbine-magnetic coupling-generator rotor train has enough to separation margin regardless of connection stiffness of magnetic coupling. The analytical FE model is guaranteed by impact test on the prototype and condition monitoring such as measurements of vibration and bearing temperature is also performed.