• 한국기계가공학회지
• /
• 제12권5호
• /
• pp.23-29
• /
• 2013

Recent advances in printing technology have increased the productivity of the roll-to-roll (R2R) printing process for printed circuitry. In the R2R printed electronics, characteristics of printed and coated layers are one of the most important issues that determine the functional quality of final products. The slot-die technology can coat a large area with high uniformity using low-viscosity materials; determining the process parameters is important to obtain excellent coating qualities. In this study, a viscocapillary model was used to predict qualities of coated layers and patterns. On the basis of analysis results, a novel meta model was derived using design of experiment methodology to improve accuracy. Sensitivity analysis was performed to define major parameters in R2R slot-die coating process. The coating speed was found to most significantly affect the coated layer thickness and was easily controlled. The performance of the proposed model is verified through experimental studies. Based on the statistical analysis results, R2R slot die process can be optimized to guarantee a desired thickness.

• Membrane and Water Treatment
• /
• 제8권4호
• /
• pp.323-336
• /
• 2017

The main challenge in many applications of acetic acid is acid dehydration and its recovery from wastewater streams. Therefore, the performance of polyamide thin film composite is evaluated to separate acetic acid from water. To reach this goal, the formation of polyamide layer on polysulfone support membrane was investigated via interfacial polymerization (IP) of meta-phenylenediamine (MPD) in water with trimesoyl chloride (TMC) in hexane. Also, the effect of synthesis conditions, such as concentration of monomers and curing temperature on separation of acetic acid from water were investigated by reverse osmosis process. Moreover, the separation mechanism was discussed. The solute permeation was carried out under applied pressure of 5 bar at $25^{\circ}C$. Surface properties of TFC membrane were characterized by ATR-FTIR, SEM and AFM. The performance test indicated that 3.5 wt% of MPD, 0.35 wt% of TMC and curing temperature of $75^{\circ}C$ are the optimum conditions. Moreover, the permeate flux was $4.3{\frac{L}{m^2\;h}}$ and acetic acid rejection was about 43% at these conditions.

• 산업경영시스템학회지
• /
• 제39권3호
• /
• pp.83-89
• /
• 2016

The group formation problem of the machine and part is a critical issue in the planning stage of cellular manufacturing systems. The machine-part grouping with alternative process plans means to form machine-part groupings in which a part may be processed not only by a specific process but by many alternative processes. For this problem, this study presents an algorithm based on self organizing neural networks, so called SOM (Self Organizing feature Map). The SOM, a special type of neural networks is an intelligent tool for grouping machines and parts in group formation problem of the machine and part. SOM can learn from complex, multi-dimensional data and transform them into visually decipherable clusters. In the proposed algorithm, output layer in SOM network had been set as one-dimensional structure and the number of output node has been set sufficiently large in order to spread out the input vectors in the order of similarity. In the first stage of the proposed algorithm, SOM has been applied twice to form an initial machine-process group. In the second stage, grouping efficacy is considered to transform the initial machine-process group into a final machine-process group and a final machine-part group. The proposed algorithm was tested on well-known machine-part grouping problems with alternative process plans. The results of this computational study demonstrate the superiority of the proposed algorithm. The proposed algorithm can be easily applied to the group formation problem compared to other meta-heuristic based algorithms. In addition, it can be used to solve large-scale group formation problems.

• 한국진공학회:학술대회논문집
• /
• 한국진공학회 2015년도 제49회 하계 정기학술대회 초록집
• /
• pp.235.1-235.1
• /
• 2015

Artificially-engineered materials, whose electromagnetic properties are not available in nature, such as negative reflective index, are called metamaterials (MMs). Although many scientists have investigated MMs for negative-reflective-index properties at the beginning, their interests have been extended to many other fields comprising perfect lenses. Among various kinds of MMs, metamaterial absorbers (MM-As) mimic the blackbody through minimizing transmission and reflection. In order to maximize absorption, the real and the imaginary parts of the permittivity and permeability of MM-As should be adjusted to possess the same impedance as that of free space. We propose a dual-wide-band and polarization-independent MM-A. It is basically a triple-layer structure made of metal/dielectric multilayered truncated cones. The multilayered truncated cones are periodically arranged and play a role of meta-atoms. We realize not only a wide-band absorption, which utilizes the fundamental magnetic resonances, but also another wide-band absorption in the high-frequency range based on the third-harmonic resonances, in both simulation and experiment. In simulation, the absorption bands with absorption higher than 90% are 3.93 - 6.05 GHz and 11.64 - 14.55 GHz, while the experimental absorption bands are in 3.88 - 6.08 GHz and 9.95 - 13.84 GHz. The physical origins of these absorption bands are elucidated. Additionally, it is also polarization-independent because of its circularly symmetric structures. Our design is scalable to smaller size for the infrared and the visible ranges.

• 센서학회지
• /
• 제12권6호
• /
• pp.289-297
• /
• 2003

본 논문에서는 정신적 과제수행 동안 EEG 뇌파의 정확한 분류방법에 관하여 기술한다. 피험자는 실험 task에서 시각적 자극에 대한 반응, 문제의 해석, 손동작 제어와 키 선택을 수행한다. 선택시간을 감지하기 위하여 측정한 뇌파로부터 $\alpha$, $\beta$, $\theta$, $\gamma$를 분리하고 4가지의 특징들을 해석한파. 이 특징들을 분석하여 각 피험자별로 공통적인 특징플로 구성된 일반 규칙을 설정한다. 본 시스템의 신경망은 1개의 은닉층을 갖는 3층의 피드포워드 신경망 구조를 가지며 학습에는 역전파 학습 알고리즘을 이용하였다. 4명의 피험자를 대상으로 설정한 알고리즘들을 적용하여 평균 87% 분류 성공률을 보였다. 본 논문에서 제안한 방법은 인지적인 정신과제 판별을 위한 방법들과 결합하여 BCI 기술을 위한 기반 기술로 활용될 수 있다.

• 한국분말재료학회지
• /
• 제12권3호
• /
• pp.186-191
• /
• 2005

A formation of aluminum hydroxide by hydrolysis of nano and micro aluminum powder has been studied. The nano aluminum powder of 80 to 100 nm in diameter was fabricated by a pulsed wire evaporation (PWE) method. The micro powder was commercial product with more than $10\;{\mu}m$ in diameter. The hydroxide type and morphology depending on size of the aluminum powder were examined by several analyses such as XRD, TEM, and BET. The hydrolysis procedure of micro aluminum powder was different from that of nano aluminum powder. The nano aluminum powder after immersing in the water was transformed rapidly to a nano fibrous boehmite, accompanying with a remarkable temperature increase, and then further transformed slowly to a stable bayerite. However, the micro powder was changed to the stable bayerite slowly and directly. The formation of fibrous aluminum hydroxide from nano aluminum powder might be due to the fine cracks which were formed by hydrogen gas pressure on the surface hydroxide layer during hydrolysis. The nano powder with large specific surface area and small size reacted more actively and faster than the micro powder, and transformed to meta-stable hydroxide in relatively short reaction time. Therefore, the formation of fibrous boehmite is special characteristic of hydrolysis of nano aluminum powder.

• Archives of Pharmacal Research
• /
• 제27권2호
• /
• pp.217-224
• /
• 2004

Previously, we reported that water-extracted Acanthopanax senticasus exhibited anti-meta-static activity by stimulating the immune system. In this study, we fractionated glycoproteins (EN-SP) from the soluble protein layer (GF-AS) of A. senticasus and determined their basic chemical properties. We also investigated the anti-tumor and immunostimulating activities of the fractionated glycoprotein, EN-SP. We found that intravenous (i.v.) administration of GF-AS dramatically inhibited metastasis of colon26-M3.1 carcinoma cells to the lung in a dose-dependent manner. In vitro analysis showed GF-AS to enhance the proliferation of splenocytes. GF-AS also stimulated peritoneal macrophage, which was followed by the production of various cytokines such as IL-1$\beta$, TNF-$\alpha$, IL-12 and IFN-${\gamma}$. Furthermore, the production of these cytokines was partially blocked when peritoneal macrophage was cultured with the polyclonal antibodies against GF-AS. The depletion of NK cells by rabbit anti-asialo GM1 serum partly abolished the inhibitory effect of GF-AS on lung metastasis of colon26-M3.1 cells. Using gel filtration, EN-SP, an active glycoprotein fraction, is isolated from GF-AS. While both GF-AS and EN-SP stimulated the proliferatation of splenocytes of normal mice, EN-SP showed higher anti-metastatic activity and more potently stimulated the proliferation of splenocytes compared to GF-AS. These results suggest the use of EN-SP, the fractionated glycoprotein from A. senticasus, can be used as a therapeutical reagent to prevent or inhibit tumor metastasis.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2014년도 추계학술대회
• /
• pp.342-345
• /
• 2014

기계학습을 이용한 네트워크 침입탐지시스템은 선택된 특징 조합에 따라 정확성 및 효율성 측면에서 크게 영향을 받는다. 하지만 일반적으로 사용되는 침입탐지용 특징들로부터 최적의 조합을 찾아내는 일은 많은 계산량을 요구한다. 예를 들어 n개로 구성된 특징들로부터 가능한 특징조합은 $2^n-1$ 개이다. 본 논문에서는 이러한 문제를 해결하기 위한 최적 특징 선택 알고리즘을 제시한다. 제안한 알고리즘은 최적화 문제 해결을 위한 대표적인 메타 휴리스틱 알고리즘인 지역탐색 알고리즘에 기반 한다. 또한 특징 조합을 평가를 위해 선택된 특징 요소와 k-means 군집화 알고리즘을 이용해 구해진 군집화의 정확성을 비용함수로 사용한다. 제안한 특징 선택 알고리즘의 평가를 위해 NSL-KDD 데이터와 인공 신경망을 사용해 특징 모두를 사용한 경우와 비교한다.

• Nuclear Engineering and Technology
• /
• 제55권9호
• /
• pp.3423-3440
• /
• 2023

In this work, a multivariate time-series machine learning meta-model is developed to predict the transient response of a typical nuclear power plant (NPP) undergoing a steam generator tube rupture (SGTR). The model employs Recurrent Neural Networks (RNNs), including the Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and a hybrid CNN-LSTM model. To address the uncertainty inherent in such predictions, a Bayesian Neural Network (BNN) was implemented. The models were trained using a database generated by the Best Estimate Plus Uncertainty (BEPU) methodology; coupling the thermal hydraulics code, RELAP5/SCDAP/MOD3.4 to the statistical tool, DAKOTA, to predict the variation in system response under various operational and phenomenological uncertainties. The RNN models successfully captures the underlying characteristics of the data with reasonable accuracy, and the BNN-LSTM approach offers an additional layer of insight into the level of uncertainty associated with the predictions. The results demonstrate that LSTM outperforms GRU, while the hybrid CNN-LSTM model is computationally the most efficient. This study aims to gain a better understanding of the capabilities and limitations of machine learning models in the context of nuclear safety. By expanding the application of ML models to more severe accident scenarios, where operators are under extreme stress and prone to errors, ML models can provide valuable support and act as expert systems to assist in decision-making while minimizing the chances of human error.

• 한국전자거래학회지
• /
• 제27권1호
• /
• pp.15-28
• /
• 2022

최근 의류 등의 특정 쇼핑몰의 온라인 시장이 크게 확대되면서, 사용자의 리뷰를 활용하는 것이 주요한 마케팅 방안이 되었다. 이를 이용한 감성분석에 대한 연구들도 많이 진행되고 있다. 감성분석은 사용자의 리뷰를 긍정과 부정 그리고 필요에 따라서 중립으로 분류하는 방법이다. 이 방법은 크게 머신러닝 기반의 감성분석과 사전기반의 감성분석으로 나눌 수 있다. 머신러닝 기반의 감성분석은 사용자의 리뷰 데이터와 그에 대응하는 감성 라벨을 이용해서 분류 모델을 학습하는 방법이다. 감성분석 분야의 연구가 발전하면서 리뷰와 함께 제공되는 이미지나 영상 데이터 등을 함께 고려하여 학습하는 멀티모달 방식의 모델들이 연구되고 있다. 리뷰 데이터에서 제품의 카테고리와 사용자별로 사용되는 단어 등의 특징이 다르다. 따라서 본 논문에서는 리뷰데이터와 제품 정보를 동시에 고려하여 감성분석을 진행한다. 리뷰를 분류하는 모델로는 기본 순환신경망 구조에서 Gate 방식을 도입한 Gated Recurrent Unit(GRU), Long Short-Term Memory(LSTM) 그리고 Self Attention 기반의 Multi-head Attention 모델, Bidirectional Encoder Representation from Transformer(BERT)를 사용해서 각각 성능을 비교하였다. 제품 정보는 모두 동일한 Multi-Layer Perceptron(MLP) 모델을 이용하였다. 본 논문에서는 사용자 리뷰를 활용한 Baseline Classifier의 정보와 제품 정보를 활용한 MLP모델의 결과를 결합하는 방법을 제안하며 실제 데이터를 통해 성능의 우수함을 보인다.