• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.30 no.3
• /
• pp.181-189
• /
• 2019

In this study, a triple-band antenna that can be used in WLAN(Wireless Local Area Network) at 2.4 GHz, 5.8 GHz, and 5G at 3.5 GHz is fabricated. The proposed antenna uses a parasitic element to show the triple band, and the reflector is used at a distance of ${\lambda}/4$ from the antenna to reduce the Specific Absorption Rate(SAR). Its dimensions are $100{\times}75{\times}1.6mm^3$ and each parameter value is optimized for better performance and a lower SAR value. As a result, we obtained a bandwidth of 540 MHz(2.02~2.56 GHz), 390 MHz(3.39~3.78 GHz), and 1,210 MHz(5.56~6.77 GHz) based on the reflection loss factor of -10 dB. In addition, the SAR values of the antenna with reflector are observed to reduce below the SAR value of international standard.

• Smart Media Journal
• /
• v.8 no.2
• /
• pp.58-65
• /
• 2019

Automatic text summarization is a process of shortening a text document by either extraction or abstraction. The abstraction approach inspired by deep learning methods scaling to a large amount of document is applied in recent work. Abstractive text summarization involves utilizing pre-generated word embedding information. Low-frequent but salient words such as terminologies are seldom included to dictionaries, that are so called, out-of-vocabulary(OOV) problems. OOV deteriorates the performance of Encoder-Decoder model in neural network. In order to address OOV words in abstractive text summarization, we propose a copy mechanism to facilitate copying new words in the target document and generating summary sentences. Different from the previous studies, the proposed approach combines accurate pointing information and selective copy mechanism based on bidirectional RNN and bidirectional LSTM. In addition, neural network gate model to estimate the generation probability and the loss function to optimize the entire abstraction model has been applied. The dataset has been constructed from the collection of abstractions and titles of journal articles. Experimental results demonstrate that both ROUGE-1 (based on word recall) and ROUGE-L (employed longest common subsequence) of the proposed Encoding-Decoding model have been improved to 47.01 and 29.55, respectively.

• Journal of Korean Society of Forest Science
• /
• v.108 no.3
• /
• pp.357-363
• /
• 2019

This study was conducted to develop a stem volume table for Quercus acutissima in Korea by using Kozak's stem taper equation. In total, 2700 tree samples were collected around the country, and growth performance was investigated through compiling data on diameters by stem height and stem analysis. In order to test the stem taper equation's fitness, the fitness index (FI), bias, and mean absolute deviation (MAD) were analyzed. The fitness of the equation was estimated at 97%, bias as 0.017, and MAD turned out to be 1.118, respectively. Furthermore, there was a statistically significant volume difference between the current volume table and the new volume table (p = 0.0008, <0.005). The result indicates that using the new volume table that reflects the actual forest will reduce the loss when assessing wood resources and will improve the accuracy of forest statistics for national and local governments. A stem volume table, the main result of this research, which is utilized in the estimated stem taper equation, will provide growth information for Quercus acutissima, one of the main broadleaf species in Korea, and will function as a management indicator for rational forest management.

• Journal of IKEEE
• /
• v.23 no.3
• /
• pp.1096-1099
• /
• 2019

In this paper, we propose the protocol for construction of smart factory. The proposed protocol for construction of smart factory consists of an OPC UA Server/Client, a technology of TSN realtime communication, a NTP & PTP time synchronization protocol, a FieldBus protocol and conversion module, a technology of saving data for data transmit latency and synchronization protocol. OPC UA server/client is a system integration protocol which makes interface industrial hardware device and supports standardization which allows in all around area and also in not independent from any platform. A technology of TSN realtime communication provides an high sensitive time management and control technology in a way of sharing specific time between devices in the field of high speed network. NTP & PTP time synchronization protocol supports IEEE1588 standardization. A fieldbus protocol and conversion module provide an extendable connectivity by converting industrial protocol to OPC. A technology of saving data for data transmit latency and synchronization protocol provide a resolution function for a loss and latency of data. Results from testing agencies to assess the performance of proposed protocol for construction of smart factory, response time was 0.1367ms, synchronization time was 0.404ms, quantity of concurrent access was 100ea, quantity of interacting protocol was 5ea, data saving and synchronization was 1,000 nodes. It produced the same result as the world's highest level.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.12
• /
• pp.272-278
• /
• 2018

Traffic volume has been greatly increasing due to urban development and the improvement of living standards, and many complaints are being raised due to the increasing road noise. As a countermeasure against these problems, highly soundproof walls are installed on the sides of roads. However, the ability to bear wind loads is a major design requirement for soundproof walls, which contributes to the exponential increases in construction costs and restricts the height of the walls. The aim of this study is to improve the performance of soundproof walls and to dramatically reduce wind loads while maintaining excellent price competitiveness. Based on Helmholz's resonator theory, a new concept is proposed for a ventilation-type soundproofing plate that can pass through a fluid like air and reduce noise. A full-scale metal soundproofing plate was produced to satisfy the quality standards of highways by conducting a sound-pressure transmission-loss test, wind tunnel test, and material quality test. To verify the reliability, the wall was manufactured and installed, and the sound insulation effect was examined by measuring the noise over time. In the future, ventilated soundproof walls on roads could create a pleasant living environment due to the high noise-insulation effect.

• The Journal of Korean Institute of Next Generation Computing
• /
• v.15 no.5
• /
• pp.64-74
• /
• 2019

Although a non-rigid registration has high demands in clinical practice, it has a high computational complexity and it is very difficult for ensuring the accuracy and robustness of registration. This study proposes a method of applying a non-rigid registration to 3D magnetic resonance images of brain in an unsupervised learning environment by using a deep-learning network. A feature vector between two images is produced through the network by receiving both images from two different patients as inputs and it transforms the target image to match the source image by creating a displacement vector field. The network is designed based on a U-Net shape so that feature vectors that consider all global and local differences between two images can be constructed when performing the registration. As a regularization term is added to a loss function, a transformation result similar to that of a real brain movement can be obtained after the application of trilinear interpolation. This method enables a non-rigid registration with a single-pass deformation by only receiving two arbitrary images as inputs through an unsupervised learning. Therefore, it can perform faster than other non-learning-based registration methods that require iterative optimization processes. Our experiment was performed with 3D magnetic resonance images of 50 human brains, and the measurement result of the dice similarity coefficient confirmed an approximately 16% similarity improvement by using our method after the registration. It also showed a similar performance compared with the non-learning-based method, with about 10,000 times speed increase. The proposed method can be used for non-rigid registration of various kinds of medical image data.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.10
• /
• pp.397-402
• /
• 2019

Accurately predicting cancer prognosis to provide appropriate treatment strategies for patients is one of the critical challenges in bioinformatics. Many researches have suggested machine learning models to predict patients' outcomes based on their gene expression data. Gene expression data is high-dimensional numerical data containing about 17,000 genes, so traditional researches used feature selection or dimensionality reduction approaches to elevate the performance of prognostic prediction models. These approaches, however, have an issue of making it difficult for the predictive models to grasp any biological interaction between the selected genes because feature selection and model training stages are performed independently. In this paper, we propose a novel two-dimensional image formatting approach for gene expression data to achieve feature selection and prognostic prediction effectively. Node2Vec is exploited to integrate biological interaction network and gene expression data and a convolutional neural network learns the integrated two-dimensional gene expression image data and predicts cancer prognosis. We evaluated our proposed model through double cross-validation and confirmed superior prognostic prediction accuracy to traditional machine learning models based on raw gene expression data. As our proposed approach is able to improve prediction models without loss of information caused by feature selection steps, we expect this will contribute to development of personalized medicine.

• Applied Chemistry for Engineering
• /
• v.22 no.4
• /
• pp.367-375
• /
• 2011

Succinic acid ester derivatives have been used as additives for the base oil of metal working fluids and pressure working oils. In this paper, a series of succinic acid alkyl half-esters were synthesized with over 97% yields by ring-opening reaction of succinic anhydride and fatty alcohol and were soluble in 100 N base oil within 1 wt% concentration. The structures and purities of ester derivatives were confirmed by $^1H-NMR$ and FT-IR spectrum and GC analysis. Anti-corrosion properties of the esters in sea water were evaluated through ASTM D665 method and weight loss method and compared to that of succinic alkyl esters without carboxylic acid group in the molecule. As the results, anti-corrosion properties of succinic acid alkyl half-esters with carboxylic acid group were better than those of succinic acid alkyl esters without carboxylic acid group. And, Anti-corrosion properties of the esters with a shorter alkyl chain of high concentration showed better performance than those with longer alkyl chain of low concentration. Inhibition efficiency % (IE%) of the esters was over 95% in the concentration of 80 ppm and corrosion rate (CR) was below 0.3 mm/year at the same concentration. Thus, the corrosion properties of succinic acid alkyl half-esters result from the carboxylic acid groups in molecules.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.24 no.6
• /
• pp.34-41
• /
• 2020

In recent years, magnitude and frequency of earthquakes have increased in Korea. Damage to a bridge, which is one of the main infrastructures, can directly lead to considerable loss of human lives. Therefore, engineers need to evaluate the seismic fragility of the structure and prepare for the possible seismic damage. In particular, the number of aging bridges over 30 years of service increases, and thus the seismic analysis and fragility requires accounting for the aging and retrofit effects on the bridge. In this study, the nonlinear static and dynamic analyses were performed to evaluate the effects of the aging and FRP retrofit on a PSC bridge. The aging and FRP retrofit were applied to piers that dominate the response of the bridge during earthquakes. The maximum displacement of the bridge increased due to the aging of the pier but decreased when FRP retrofit applied to the aged pier. In addition, seismic fragility analysis was performed to evaluate the seismic behavior of the bridge combined with the seismic performance of the pier. Compared with the aged bridge, the FRP retrofit bridge showed a decrease in the seismic fragility in all levels of damage. The reduction of the seismic fragility in the FRP bridge was prominent as the value of PGA and level of damage increased.

• Journal of the Korean Society of Radiology
• /
• v.15 no.4
• /
• pp.547-554
• /
• 2021

In order to quickly and accurately diagnose pneumonia on a chest X-ray image, different batch sizes of 4, 8, 16, and 32 were applied to the same Xception deep learning model, and modeling was performed 3 times, respectively. As a result of the performance evaluation of deep learning modeling, in the case of modeling to which batch size 32 was applied, the results of accuracy, loss function value, mean square error, and learning time per epoch showed the best results. And in the accuracy evaluation of the Test Metric, the modeling applied with batch size 8 showed the best results, and the precision evaluation showed excellent results in all batch sizes. In the recall evaluation, modeling applied with batch size 16 showed the best results, and for F1-score, modeling applied with batch size 16 showed the best results. And the AUC score evaluation was the same for all batch sizes. Based on these results, deep learning modeling with batch size 32 showed high accuracy, stable artificial neural network learning, and excellent speed. It is thought that accurate and rapid lesion detection will be possible if a batch size of 32 is applied in an automatic diagnosis study for feature extraction and classification of pneumonia in chest X-ray images using deep learning in the future.