http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
전유택(Yootaek JEON),박승환(Seunghwan PARK),박병철(Byungcheol PARK),강성균(Sunggyun KANG) 대한기계학회 2017 대한기계학회 춘추학술대회 Vol.2017 No.11
As interest in sustainable and eco-friendly energy increases, demand for hydrogen energy is also increasing. Hydrogenogenic CO oxidation (CO + H₂O → CO₂ + H₂) has the potential for H₂ production as a clean renewable fuel. In this study, it was developed economic biological hydrogen technology utilizing renewable resources such as industrial by-product gas including carbon monoxide, using the Thermococcus Onnurineus NA1. Pilot plant was designed and constructed in order to convert from CO in LDG which is a by-product gas of steel mill to H₂ and implemented the technology through demonstration. 2.4 Nm³/h of LDG was continuously supplied to 1t culture medium reaction tank. As a result, 97% CO gas in LDG was converted into hydrogen gas and hydrogen content in the product gas is 40% or more. The purity of product gas was increased to 99.9% hydrogen gas by PSA. The high purity hydrogen gas was successively used to generate electricity by fuel cell system which has developed in Hyundai Steel. This results is the first demonstration of H₂ production from steel mill by-product gas by a carboxydotrophic hydrogenogenic microbe.
전유택(Youtaek Jeon),조형준(HyungJun Cho) 한국데이터정보과학회 2019 한국데이터정보과학회지 Vol.30 No.3
의사결정나무는 주어진 예측 변수 공간을 여러 개의 단순한 사각 형태의 공간으로 분할해가면서목표 대상을 분류하거나 예측할 수 있다. 의사결정나무는 사용이 용이하며, 이해하고 설명하기 쉬운 구조를 지니고 있다. 하지만 서포트 벡터 머신과 같이 수직선 혹은 수평선 형태와 비교해 훨씬 복잡한 결정 경계를 지닌 분석 기법에 비해 예측 정확도가 상대적으로 낮다는 단점을 가지고 있다. 본 연구에서는 유연한 결정 경계를 가지는 의사결정나무 방법론을 제안한다. 뿌리 마디에서부터 유연한 결정 경계를 지니는 기법을 적용하여 예측 정확도를 향상시킨다. 또한, 이는 나무 모형의 크기도 줄여준다. Decision trees can categorize or predict target objects by dividing given predictor spaces into several simple rectangular spaces. Decision-making trees have the advantages of being easy to use, understand and explain. However, it tends to have relatively lower predictive accuracy than analysis methods with more complex decision boundaries, such as support vector machines. In this study, a decision tree methodology with flexible decision boundaries is proposed. The model based hybrid decision tree improves prediction accuracy by fitting a flexible decision boundary to each node. Furthermore, it reduces a tree size significantly.
전유택(Jeon, Yoo-Taek),정경우(Chung, Kyeong-Woo),나상묵(Na, Sang-Mook) 한국신재생에너지학회 2005 한국신재생에너지학회 학술대회논문집 Vol.2005 No.06
Bipolar plate is the main part with MEA in automotive PEMFC. It must have a good electrical conductivity and excellent corrosion resistance, be cost effective. Therefore, stainless steels have been studied by many researchers because of its corrosion resistance and cost benefits. But their properties are not sufficient for the application to bipolar plate for automotive PEMFC. In this work, we have performed stamping using various commercial stainless steels to select candidate material for biploar plate and to derive design parameters for stamping simulation. The results showed that a small curvature at the corner of flow field is more favorable due to easier a plastic deformation. Stamping process was simulated by changing surface condition, and the size and angle of channel. The optimum shape and spring back phenomena were evaluated. Surface coating was applied to increase the corrosion resistance and electrical conductivity of stainless steel. The electrical interfacial resistance was 10 to 15m{Omega}cm² under clamping force of 150psi. But corrosion resistance of coating on the stainless steel was not good due to the unstableness of microstructure.