• Proceedings of the Korea Concrete Institute Conference
• /
• 2003.05a
• /
• pp.961-966
• /
• 2003

In recent, the design of diaphragm which is representative disturbed region in PSC box girder bridge have been performed according to the empirical method or beam theory. But, these methods couldn't be described the behavior of the end diaphragm, and placed reinforcements accurately. As the compressive stress transferred by the web concentrated on the lower parts of diaphragm, it was demonstrated that the basic assumption of 2-D strut-and-tie model for the diaphragm that the compressive stress acts on the upper parts of the diaphragm is wrong. Meanwhile, in this research, after analyzing the variables of end diaphragm, the 2-D strut-and-tie models appropriate to each cases are proposed. And, the problems of 2-D strut-and-tie model were analyzed, so 3-D strut-and-tie model is proposed as well. There is no codes which include the demonstration of safety of 3-D strut-and-tie model. Hence, for nodes, the stresses at the elements which included the singular node in strut-and-tie model were investigated using the finite element analysis. And, the stress states of strut has one direction, so effective stresses were considered at the stage, dimensioning of the model. From the results, 3-D strut-and-tie model could predict the behavior of end diaphragm accurately, and design of reinforcement could be performed economically.

• The Transactions of the Korean Institute of Electrical Engineers B
• /
• v.55 no.9
• /
• pp.459-464
• /
• 2006

The detent force by end-effect has an undesired influence on moving coil type Permanent Magnet Linear Synchronous Motor(PMLSM). So, the reduction of detent force by end-effect is especially required for the improvement of thrust characteristics. In this paper, in order to reduce detent force by end-effect, the auxiliary-teeth is installed at the end part of mover. It is also analyzed by Finite Element Analysis(FEA) and optimized by using neural network. By comparison, the detent force is reduced about 41.4[%] comparing to that of basic model.

• Journal of Korea Technical Association of The Pulp and Paper Industry
• /
• v.35 no.5
• /
• pp.26-36
• /
• 2003

The wet-end dynamics of a paper mill was analyzed to characterize its dynamic behavior during the grade change of paper. The model representing the wet-end section is developed based on the mass balance relationships written for the simplified wet-end white water network. From the linearization of dynamic model, higher-order Laplace transfer functions were obtained followed by the reduction procedure to give simple lower-order models in the form of 1$^{st}$-order or 2$^{nd}$ -order plus dead times. The dynamic response of the wet-end is influenced both by the white water volume and by the level of wire retention. Effects of key manipulated variables such as the thick stock flow rate, the ash flow rate and the retention aid flow rate on the major controlled variables were analyzed by numerical simulations. The simple dynamic model developed in the present study can be effectively used in the operation and control of paper mills.s.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.15 no.2
• /
• pp.51-60
• /
• 2020

Optical character recognition (OCR) has been studied for decades because it is very useful in a variety of places. Nowadays, OCR's performance has improved significantly due to outstanding deep learning technology. Thus, there is an increasing demand for commercial-grade but affordable OCR systems. We have developed a low-cost, high-performance OCR system for the industry with the cheapest embedded developer kit that supports GPU acceleration. To achieve high accuracy for industrial use on limited computing resources, we chose a state-of-the-art text recognition algorithm that uses an end-to-end deep learning network as a baseline model. The model was then improved by replacing the feature extraction network with the best one suited to our conditions. Among the various candidate networks, EfficientNet-B3 has shown the best performance: excellent recognition accuracy with relatively low memory consumption. Besides, we have optimized the model written in TensorFlow's Python API using TensorFlow-TensorRT integration and TensorFlow's C++ API, respectively.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.24 no.7 s.178
• /
• pp.1681-1688
• /
• 2000

A dynamic model for the prediction of surface topography in high speed end milling process is developed. In this model the effect of tool runout, tool deflection and spindle vibration were taken in to account. An equivalent diameter of end mill is obtained by finite element method and tool deflection experiment. A modal parameter of machine tool is extracted by using frequency response function. The tool deflection, spindle vibration chip thickness and cutting force were calculated in dynamic cutting condition. The tooth pass is calculated at the current angular position for each point of contact between the tool and the workpiece. The new dynamic model for surface predition are compared with several investigated model. It is shown that new dynamic model is more effective to predict surface topography than other suggested models. In high speed end milling, the tool vibration has more effect on surface topography than the tool deflection.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1997.04a
• /
• pp.788-793
• /
• 1997

The surface,generated by end milling operation, is deteriorated by tool runout,vibration,friction,tool deflection, etc. In many source,deflection of tool affects to surfave accuracy. To develop a surface accracy model,method for the prediction of the topography of machined surfaces has been developed based on models of machine tool kinematics and cutting tool geometry. This model accounts for not only the ideal geometrical surface, but also the deflection of tool resulted in cutting force. For the more accurate prediction of cutting force,flexible end mill model is used to simulate cutting process. Compute simu;ation have shown the feasibility of the surface generation system.

• Proceedings of The KACE
• /
• 2017.08a
• /
• pp.57-59
• /
• 2017

대화 시스템(dialogue system)은 텍스트나 음성을 통해 다양한 분야에서 특정한 목적을 수행할 수 있는 시스템이다. 대화 시스템을 구현하기 위한 방법으로 인공 신경망(neural network)을 기반으로한 end-to-end learning 방식이 제안되었다. End-to-end learning 방식을 이용한 식당 예약 시스템 모델의 학습을 위해 페이스북은 영어로 이루어진 식당 예약에 관련된 학습 대화 데이터셋(The 6 dialog bAbI tasks)을 구축하였다. 하지만 end-to-end learning 방식의 학습은 많은 학습 데이터가 필요하다는 단점이 존재하는데, 액션 템플릿(action template)의 정의를 통해 도메인 지식을 표현함으로써 일반적인 end-to-end learning 방식보다 적은 학습량으로 좋은 성능의 모델을 학습할 수 있는 Hybrid Code Network 구조를 제안한 연구가 있다. 본 논문에서는 Hybrid Code Network 구조를 이용하여 한국어 식당 예약 시스템을 구축할 수 있는 방법을 제안하고, 한국어로 이루어진 식당 예약에 관련한 학습 대화 데이터를 구축하는 방법을 제안한다.

• Geomechanics and Engineering
• /
• v.23 no.6
• /
• pp.523-533
• /
• 2020

The current study focuses on the effect of the end shape of steel pipe piles on installation effort and bearing resistance using the pressing method of installation under dense ground conditions. The effect of pile rotation on the installation effort and bearing resistance is also investigated. The model steel piles with a flat end, cone end and cutting-edge end were used in this study. The test results indicated that cone end pile with the pressing method of installation required the least installation effort (load) and showed higher ultimate resistance than flat and cutting-edge end piles. However, pressing and rotation during cutting-edge end pile installation considerably reduces the installation effort (load and torque) if pile penetration in one rotation equal to the cutting-edge depth. Inclusion of rotation during pile installation reduces the ultimate bearing resistance. However, if penetration of the cutting-edge end pile equal to the cutting-edge depth in one rotation, the reduction in ultimate resistance can be minimized. In comparing the cone and cutting-edge end piles installed with pressing and rotation, the least installation effort is observed in the cutting-edge end pile installed with penetration rate equal to the cutting-edge depth per rotation.

• International Journal of Computer Science & Network Security
• /
• v.23 no.10
• /
• pp.107-114
• /
• 2023

The study provides the identification of vulnerabilities in the security issues by Wireless Network. To achieve it the research focus on packet flow analysis, end to end data communication, and the security challenges (Cybercrime, insider threat, attackers, hactivist, malware and Ransomware). To solve this I have used the systematic literature review mechanisms and demonstrative tool namely Wireshark network analyzer. The practical demonstration identifies the packet flow, packet length time, data flow statistics, end- to- end packet flow, reached and lost packets in the network and input/output packet statics graphs. Then, I have developed the proposed model that used to secure the Wireless network solution and prevention vulnerabilities of the network security challenges. And applying the model that used to investigate the security challenges and vulnerabilities of cloud computing services is used to fulfill the network security goals in Wireless network. Finally the research provides the model that investigate the security challenges and vulnerabilities of cloud computing services in wireless networks

• Phonetics and Speech Sciences
• /
• v.10 no.1
• /
• pp.39-48
• /
• 2018

A typical statistical parametric speech synthesis (text-to-speech, TTS) system consists of separate modules, such as a text analysis module, an acoustic modeling module, and a speech synthesis module. This causes two problems: 1) expert knowledge of each module is required, and 2) errors generated in each module accumulate passing through each module. An end-to-end TTS system could avoid such problems by synthesizing voice signals directly from an input string. In this study, we implemented an end-to-end Korean TTS system using Google's Tacotron, which is an end-to-end TTS system based on a sequence-to-sequence model with attention mechanism. We used 4392 utterances spoken by a Korean female speaker, an amount that corresponds to 37% of the dataset Google used for training Tacotron. Our system obtained mean opinion score (MOS) 2.98 and degradation mean opinion score (DMOS) 3.25. We will discuss the factors which affected training of the system. Experiments demonstrate that the post-processing network needs to be designed considering output language and input characters and that according to the amount of training data, the maximum value of n for n-grams modeled by the encoder should be small enough.