• Proceedings of the Korea Society for Energy Engineering kosee Conference
• /
• 2002.11a
• /
• pp.171-184
• /
• 2002

Approx. 200,000 bpd vacuum residue oil is produced from oil refineries in Korea. These are supplying to use asphalt, high sulfur fuel oil, and upgrading at the residue hydro-desulfurization unit. Vacuum residue oil has high energy content, however high sulfur content and high concentration of heavy metals represent improper low grade fuel. To meet growing demand for effective utilization of vacuum residue oil from refineries, recently some of the oil refinery industries in Korea, such as SK oil refinery and LG Caltex refinery, have already proceeded feasibility study to construct 435-500 MWe IGCC power plant and hydrogen production facilities. Recently, KIER(Korea Institute of Energy Research) are studing on the Vacuum Residue gasification process using an oxygen-blown entrained-flow gasifier. The experiment runs were evaluated under the reaction temperature : 1,100~1,25$0^{\circ}C$, reaction pressure : 1~6kg/$\textrm{cm}^2$G, oxygen/V.R ratio : 0.8~0.9 and steam/V.R ratio : 0.4-0.5. Experimental results show the syngas composition(CO+H$_2$) : 85~93%, syngas flow rate : 50~110Mm$^3$/hr, heating value : 2,300~3,000 ㎉/Nm$^3$, carbon conversion : 65~92, cold gas efficiency : 60~70%. Also equilibrium modeling was used to predict the vacuum residue gasification process and the predicted values were compared reasonably well with experimental data.

• Journal of Energy Engineering
• /
• v.12 no.1
• /
• pp.49-57
• /
• 2003

Approx. 200.000 bpd vacuum residue oil is produced from oil refineries in Korea, and is supplied to use asphalt, high sulfur fuel oil and for upgrading at the residue hydro-desulfurization unit. Vacuum residue oil has high energy content, however its high sulfur content and high concentration of heavy metals represent improper low grade fuel. To meet growing demand for effective utilization of vacuum residue oil from refineries, recently some of the oil refinery industries in Korea, such as SK oil refinery and LG Caltex refinery, have already proceeded feasibility study to construct 435~500 MWe IGCC power plant and hydrogen production facilities. Recently, KIER (Korea Institute of Energy Research) are studying on the Vacuum Residue gasification process using an oxygen-blown entrained-flow gasifier. The experiment runs were evaluated under the reaction temperature: 1.100~l,25$0^{\circ}C$, reaction pressure: 1~6 kg/$\textrm{cm}^2$G, oxygen/V.R ratio: 0.8~0.9 and steam/V.R ratio: 0.4~0.5. Experimental results show the syngas composition (CO+H$_2$): 85~93%, syngas flow rate: 50~l10 Nm$^3$/hr, heating value: 2,300~3,000 k㎈/Nm$^3$, carbon conversion: 65~92, cold gas efficiency: 60~70%. Also equilibrium modeling was used to predict the vacuum residue gasification process and the predicted values were compared reasonably well with experimental data.

• Journal of the Korea Concrete Institute
• /
• v.28 no.5
• /
• pp.555-562
• /
• 2016

The objective of the present study is to evaluate the shear transfer capacity of transverse reinforcement at the concrete interfaces with smooth construction joint. The transverse reinforcing bars were classified into two groups: V-type for the arrangement perpendicular to the interface and X-type for inclined-crossing arrangement. The transverse reinforcement ratio at the interface varied from 0.0045 to 0.0135 for V-type and 0.0064 to 0.0045 for X-type. The mechanism analysis proposed for monolithic concrete interface, derived based on the upper-bound theorem of concrete plasticity, was modified to evaluate the shear friction capacity of concrete interfaces with smooth construction joint. Test results showed that the specimens with X-type reinforcement had lower amount of relative slippage at the interface and higher shear friction capacity than the companion specimens with V-type reinforcement. This observation was independent of the unit weight of concrete. The mean and standard deviation of the ratios between the experimental shear friction strength of smooth construction joints and predictions obtained from the proposed model are 1.07 and 0.14, respectively.

• Journal of the Korean Chemical Society
• /
• v.31 no.5
• /
• pp.406-412
• /
• 1987

The electrical conductivities of ${\alpha}-Fe_2O_3$ containing 2.5 and 5.0mol% of cadmium were measured from 300 to $900^{\circ}C$ under oxygen pressures of 10$^{-7}$ to 10$^{-1}$ atmosphere. Plots of log ${\sigma}$ vs. 10$^3$/T show the extrinsic conductivity at oxygen pressure higher than $5{\times}10$^{-2}$atm. The transition points appear at about 550$^{\circ}$C and the activation energies are 1.34 eV for the intrinsic region and 0.50 eV for the extrinsic region on 5mol% Cd-doped ${\alpha}-Fe_2O_3$. The extrinsic conductivity disappears at oxygen partial pressures lower than $5{\times}10$^{-2}$ atm, and the intrinsic conductivity predominates. The electrical conductivities decrease with increasing mol% of cadmium doped. The predominant defect of ${\alpha}-Fe_2O_3$ doped with Cd is believed to be Fe${2+}$ interstitial for the intrinsic, however, oxygen vacancy predominates for the extrinsic region. The electrical conduction mechanisms are proposed and the conduction band model is suggested for the extrinsic region.

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.5
• /
• pp.169-176
• /
• 2020

Recently, deep neural networks (DNNs) have brought revolutions to many mobile and embedded devices by providing human-level machine intelligence for various applications. However, high inference accuracy of such DNNs comes at high computational costs, and, hence, there have been significant efforts to reduce computational overheads of DNNs either by compressing off-the-shelf models or by designing a new small footprint DNN architecture tailored to resource constrained devices. One notable recent paradigm in designing small footprint DNN models is sharing parameters in several layers. However, in previous approaches, the parameter-sharing techniques have been applied to large deep networks, such as ResNet, that are known to have high redundancy. In this paper, we propose a parameter-sharing method for already parameter-efficient small networks such as ShuffleNetV2. In our approach, small templates are combined with small layer-specific parameters to generate weights. Our experiment results on ImageNet and CIFAR100 datasets show that our approach can reduce the size of parameters by 15%-35% of ShuffleNetV2 while achieving smaller drops in accuracies compared to previous parameter-sharing and pruning approaches. We further show that the proposed approach is efficient in terms of latency and energy consumption on modern embedded devices.

• Journal of Bio-Environment Control
• /
• v.19 no.3
• /
• pp.123-129
• /
• 2010

A study on modeling of medium temperature drops of the elevated-bench hydroponic system for strawberry cultivation during low temperature season was conducted. Four different conditions were used for the experiment. These consisted of two kinds of bed types (plant, V), four kinds of medium (rice, perlite, rice hulls80% and peatmoss20%, perlite80% and peatmoss20%), two kinds of mulched bed (mulched, non mulched) and four kinds of greenhouse air temperature (l.5, 3.2, 5.0, $6.7^{\circ}C$), and the results were summarized as follows: Temperature drop of medium in the V-bed was slower than that in the plant bed, showing better insulation effect of V-bed. Temperature drop of medium with mulching on the top of the bed was slower than the case without mulching, as a result, the beneficial effect of temperature drop was appeared in mulched bed. Linear regression of the temperature descent rate and the temperature difference between medium and air showed significant correlation. The regression equation for the Pearlite80% and Peatmoss20% in the V-bed was f(x) = -0.2656 + 0.1345x at the $R^2$ of 0.9269. Using the model, the temperature drop during night can be predicted for the various media at the different depths.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.5 no.4
• /
• pp.338-349
• /
• 1993

Most of numerical models for studying tide-induced residual currents (TIRC) were to dimensional depth averaged models which were confined to anlayze the horizontal structure of TIRC. In this study, TIRC occurring around a circular island was simulated with the three-dimensional spectral model which employed by the finite difference method in the horizontal direction and the expansion of basis function in the vertical direction. The main results of numerical experiment can be summarized as follows. Firstly. both topographic and nonlinear effect played an important role in the generation of TIRC. Secondly, when the currents were rotary clockwise. the horizontal structure of TIRC appeared to rotate in the same direction. These results were consistent with those of previous studies of two-dimensional numerical models.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.14 no.4
• /
• pp.80-85
• /
• 2015

The era of intelligent transportation system(ITS) has been arrived by applying information and communication technology(ICT) to the traffic. One of these technological advances is a wireless communication technology for a high speed vehicle to be connected to an infrastructure(V2I). A variety of road traffic safety services and operator comfort services are being developed by means of WAVE(Wireless Access in a Vehicular Environment) based on IEEE802.11p Standard. In this paper, the link budget is analyzed to provide a reliable quality of these ITS services. Log-distance model and two-ray model is employed for the wave propagation path loss model which is adequate for a highway environment. Reliable cell coverage is suggested for ITS services from the link budget.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2011.02a
• /
• pp.42-42
• /
• 2011

본 연구에서는 광학소자에 폭넓게 이용되는 AlGaP III-V족 화합물 반도체 중에서 한쪽 끝 이 종화합물인 AlP의 유전함수를 0.75~5.05 eV의 에너지 영역에서 타원편광분석법을 이용하여 분석하였다. AlP는 산소와 급격히 반응하기 때문에, 대기 중에서 물질 고유의 광특성이 유지되기 어려울 뿐만 아니라, 박막 위에 생성되는 산화막 때문에 순수한 AlP의 유전함수 측정이 불가능 하다. 본 연구에서는 물질의 유전함수에 미치는 산화 효과를 최소화하기 위하여 Molecular Beam Epitaxy로 성장한 $1.0{\mu}m$ 두께의 AlP 박막을 초고진공 상태의 chamber 안에서 타원편광분석기를 이용하여 실시간으로 측정하였다. 박막의 투명도에 의해 나타나는 간섭 pattern과 표면거칠기 효과로 인한 유전함수의 왜곡을 보정하기 위하여 변수화 모델이 이용되었으며 다층 변수화모델 계산을 통하여 순수한 AlP의 유전함수를 얻어낼 수 있었다. 본 연구에서 측정된 순수한 AlP의 유전함수는 타원편광분석기를 이용한 최초의 실험결과로서 이차미분을 이용한 전이점 (Critical Point) 분석결과 이론적인 electronic band structure에서 $E_1$, $E_1+_{{\Delta}_1}$, $E_2$에 해당하는 전이점들을 확인할 수 있었다.

• The Bulletin of The Korean Astronomical Society
• /
• v.36 no.1
• /
• pp.64.2-64.2
• /
• 2011

MIRIS (Multi-purpose InfraRed Imaging System), 다목적 적외선 영상시스템은 한국천문연구원에서 개발하고 있는 과학위성 3호의 주 탑재체이다. MIRIS 우주관측카메라는 한국에서 최초로 발사되는 천문우주관측용 적외선 우주망원경이다. 그 유효 구경은 80mm 이고, 탑재되는 검출기는 Teledyne사의 PICNIC $256{\times}256$ Array 이며, 이 검출기를 적용한 관측 화각(FoV)은 $3.67 deg{\times}3.67 deg$, Pixel Scale은 51.6 arcsec/pixel 이다. MIRIS는 현재 비행모델의 납품을 앞두고 우주환경 시험을 준비 및 진행하고 있으며, 시스템의 최적화 작업을 함께 수행하고 있다. 최근에 과학기술위성 3호의 발사체가 러시아 Dnepr로 결정되면서 시험 조건이 변경된 시험 항목에 대해서 EQM의 Sine진동, 충격 시험이 Qualification level로 진행되었다. 그리고 MIRIS 비행모델의 열진공 환경 시험 및 진동시험에 대한 준비 현황을 보고한다. MIRIS 비행 모델의 환경시험은 실제 위성이 겪는 acceptance level로 진행되며, 모든 시험을 통과하면 최종 납품이 이뤄질 예정이다. 또한 시스템의 최적화를 위해 수행했던 조립의 수정 항목들도 함께 보고한다.