• Proceedings of the Korea Water Resources Association Conference
• /
• 2019.05a
• /
• pp.203-203
• /
• 2019

도시화로 인한 불투수면의 증가는 물순환 왜곡, 비점오염원 발생 및 수생태계 건상성 훼손 등을 야기시키며, 이를 해결하기 위하여 다양한 LID 기법을 적용하고 있다. 일반적으로 LID 내 적용 되는 여재들은 무기성여재로 중량이 크고 미세공극의 부재로 물리화학적 및 생물학적 저감기능이 제한적이다. 따라서 본 연구에서는 중량성이 낮은 생물폐자원을 선정하여 LID 시설 적용성평가를 수행하고자 한다. 생물폐자원은 발생량, 경량성 및 용이성을 고려하여 꼬막껍질(CS)과 호두껍질(WS)을 선정하였다. 생물폐자원의 산화부식을 고려하여 무기성 여재인 화산석과 혼합하여 Bioretention 시설에 적용하였으며, 여재 혼합비율에 따라 총 3가지의 Case 로 구성하였다. 식생은 구절초와 꽃댕강나무를 식재하였으며, 여재의 물리적 특성 분석을 위하여 적용 전과 후의 SEM(Scanning Electronic Microscope)을 수행하였다. 모니터링은 도로퇴적물 100g과 물 110L를 제조하여 인공강우유출수를 이용하여 수행하였으며, $0.0003{\sim}0.007m^3/sec$의 유속으로 주입하였다. 시설의 유입 및 유출부에서 유량 측정 및 수질 시료를 채취하였으며, 채취된 시료는 수질오염공정 시험법에 준하여 입자상 물질, 유기물, 영양물질 및 중금속 등을 분석하였다. Bioretention 시설의 모니터링 결과를 이용하여 물수지 및 TSS 저감 효율을 산정하였으며. 물수지 분석결과 시설의 저류율은 Case 1(soil) > Case 3(WS+RV) > Case 2(CS+RV) 순으로 나타났다. 시설 내 공극률이 가장 낮았던 Case 1에서 저류율이 약 55%로 가장 높게 것으로 분석되었다. Case 3(WS+RV)은 Case 2(WS+RV)와 시설 내 공극률이 유사함에도 불구하고 저류율이 약 10% 높은 것으로 분석되었다. 오염물질 저감효율 분석 결과, TSS와 TP의 제거효율은 모든 Case에서 약 75% 이상으로 높게 나타났으며, COD의 경우 모래를 적용한 Case 1에 비해 생물폐자원인 꼬막껍질과 호두껍질을 적용한 Case에서 약 1.3배 이상 높게 나타났다. 호두껍질과 꼬막껍질의 SEM 분석 결과 표면에 다공성이 형성되어 있는 것으로 조사되었다. 이는 여과 및 저류기작으로 인한 물순환 효과증대와 다공성과 돌기사이로 인한 입자상의 물질 여과 및 흡착으로 인하여 오염물질의 제거효율이 증대 된 것으로 평가된다. LID시설 내 생물폐자원과 무기성여재를 적절히 배합하여 복합여재로 조성할 경우 침하현상을 방지할 뿐만 아니라 저류 및 침투기능 향상과 미생물의 서식환경을 제공하기에 물순환 회복 및 비점오염물질 저감에 기여할 것으로 평가된다.

• Journal of Chest Surgery
• /
• v.37 no.3
• /
• pp.201-209
• /
• 2004

Extracorporeal life support (ECLS) system is a device for respiratory and/or heart failure treatment, and there have been many trials for development and clinical application in the world. Currently, a non-pulsatile blood pump is a standard for ECLS system. Although a pulsatile blood pump is advantageous in physiologic aspects, high pressure generated in the circuits and resultant blood cell trauma remain major concerns which make one reluctant to use a pulsatile blood pump in artificial lung circuits containing a membrane oxygenator. The study was designed to evaluate the hypothesis that placement of a pressure-relieving compliance chamber between a pulsatile pump and a membrane oxygenator might reduce the above mentioned side effects while providing physiologic pulsatile blood flow. The study was performed in a canine model of oleic acid induced acute lung injury (N=16). The animals were divided into three groups according to the type of pump used and the presence of the compliance chamber, In group 1, a non-pulsatile centrifugal pump was used as a control (n=6). In group 2 (n=4), a single-pulsatile pump was used. In group 3 (n=6), a single-pulsatile pump equipped with a compliance chamber was used. The experimental model was a partial bypass between the right atrium and the aorta at a pump flow of 1.8∼2 L/min for 2 hours. The observed parameters were focused on hemodynamic changes, intra-circuit pressure, laboratory studies for blood profile, and the effect on blood cell trauma. In hemodynamics, the pulsatile group II & III generated higher arterial pulse pressure (47$\pm$ 10 and 41 $\pm$ 9 mmHg) than the nonpulsatile group 1 (17 $\pm$ 7 mmHg, p<0.001). The intra-circuit pressure at membrane oxygenator were 222 $\pm$ 8 mmHg in group 1, 739 $\pm$ 35 mmHg in group 2, and 470 $\pm$ 17 mmHg in group 3 (p<0.001). At 2 hour bypass, arterial oxygen partial pressures were significantly higher in the pulsatile group 2 & 3 than in the non-pulsatile group 1 (77 $\pm$ 41 mmHg in group 1, 96 $\pm$ 48 mmHg in group 2, and 97 $\pm$ 25 mmHg in group 3: p<0.05). The levels of plasma free hemoglobin which was an indicator of blood cell trauma were lowest in group 1, highest in group 2, and significantly decreased in group 3 (55.7 $\pm$ 43.3, 162.8 $\pm$ 113.6, 82.5 $\pm$ 25.1 mg%, respectively; p<0.05). Other laboratory findings for blood profile were not different. The above results imply that the pulsatile blood pump is beneficial in oxygenation while deleterious in the aspects to high pressure generation in the circuits and blood cell trauma. However, when a pressure-relieving compliance chamber is applied between the pulsatile pump and a membrane oxygenator, it can significantly reduce the high circuit pressure and result in low blood cell trauma.

• Journal of Wetlands Research
• /
• v.26 no.2
• /
• pp.147-159
• /
• 2024

Lack of streamflow observations makes model calibration difficult and limits model performance improvement. Satellite-based remote sensing products offer a new alternative as they can be actively utilized to obtain hydrological data. Recently, several studies have shown that artificial intelligence-based solutions are more appropriate than traditional conceptual and physical models. In this study, a data-driven approach combining various recurrent neural networks and decision tree-based algorithms is proposed, and the utilization of satellite remote sensing information for AI training is investigated. The satellite imagery used in this study is from MODIS and SMAP. The proposed approach is validated using publicly available data from 25 watersheds. Inspired by the traditional regionalization approach, a strategy is adopted to learn one data-driven model by integrating data from all basins, and the potential of the proposed approach is evaluated by using a leave-one-out cross-validation regionalization setting to predict streamflow from different basins with one model. The GRU + Light GBM model was found to be a suitable model combination for target basins and showed good streamflow prediction performance in ungauged basins (The average model efficiency coefficient for predicting daily streamflow in 25 ungauged basins is 0.7187) except for the period when streamflow is very small. The influence of satellite remote sensing information was found to be up to 10%, with the additional application of satellite information having a greater impact on streamflow prediction during low or dry seasons than during wet or normal seasons.

• Journal of Biomedical Engineering Research
• /
• v.19 no.1
• /
• pp.47-52
• /
• 1998

To develop the supplementary heart valve which could reduce the complications of prosthetic heart valve, in vitro experiments were performed to analyze the movement of the supplement valves from the view point of the flow dynamics and leaflet dynamics. The autologous pericardium was attached to deficient portion of the porcine valve which was dissected completely and partially. The pericardium was treated by buffered glutaraldehyde solution in the preshaped mould to preserve the shape of the leaflet. The function of the valves were evaluated in the mock circulatory system and three sets of experiment were performed. The instantaneous motion of the leaflet was pictured by the 35mm camera and the pressure drop through the valve and cardiac output were measured. The supplementary valve which was replaced completely performed better thatn the valve replaced partially. In vitro experiments showed that the supplementary valve which was replaced completely performed better than the valve replaced partially. In vitro experiments showed that the supplementary valve could undertake the prosthetic heart valve in clinic.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.14 no.6
• /
• pp.1171-1180
• /
• 2019

Emotion is a psycho-physiological process that plays an important role in human interactions. Affective computing is centered on the development of human-aware artificial intelligence that can understand and regulate emotions. This field of study is also critical as mental diseases such as depression, autism, attention deficit hyperactivity disorder, and game addiction are associated with emotion. Despite the efforts in emotions recognition and emotion detection from nonstationary, detecting emotions from abnormal EEG signals requires sophisticated learning algorithms because they require a high level of abstraction. In this paper, we investigated LSTM hyperparameters for an optimal emotion EEG classification. Results of several experiments are hereby presented. From the results, optimal LSTM hyperparameter configuration was achieved.

• Journal of KIISE:Software and Applications
• /
• v.33 no.4
• /
• pp.440-448
• /
• 2006

Self-Organizing Map (SOM) is an unsupervised learning neural network and it is used for preserving the structural relationships in the data without prior knowledge. SOM has been applied in the study of complex problems such as vector quantization, combinatorial optimization, and pattern recognition. This paper proposes a new usage of SOM as a tool for schema transformation hoping to achieve more efficient genetic process. Every offspring is transformed into an isomorphic neural network with more desirable shape for genetic search. This helps genes with strong epistasis to stay close together in the chromosome. Experimental results showed considerable improvement over previous results.

• Journal of Korea Technical Association of The Pulp and Paper Industry
• /
• v.42 no.4
• /
• pp.15-24
• /
• 2010

Many efforts to understand the deterioration processes of the beeswax-treated volume of "The Annals of Joseon Dynasty" have been made. However, most previous studies have focused on individual sample sheet of the beeswax-treated paper but not book volume format. In this study, humid cycling accelerated aging for duplicated beeswax-treated book volume and Hanji book volume were carried out to examine differences in the deterioration of different parts of each volume as well as between the two book volumes during the aging. As results, it is found that the deterioration rate for the beeswax-treated volume is higher than that for the Hanji book volume. Different parts in each volume show different magnitude of deterioration. In particular, outer sides in both beeswax-treated and Hanji book volumes, which are directly exposed to the air, are deteriorated more seriously than inner sides. It is also observed that inner sides are considerably deteriorated during the aging, implying that inner deterioration may have different mechanisms from outer deterioration.

• Proceedings of the Korean Association of Geographic Inforamtion Studies Conference
• /
• 2010.06a
• /
• pp.283-284
• /
• 2010

강우가 있을 때 토양수분은 증가하여 지표온도의 상승을 억제하고, 강우와 증발산을 통해 토양수분은 대기와 지형을 연결하는 중요한 상태변수(Yoo et al., 2001)가 된다. 이에 따라 물순환의 이해와 적절한 모형의 개발을 위해서는 강우 및 토양수분의 원격측정이 필수적일 뿐 아니라 관측 격자 내에서 일어나는 변화도에 대한 이해가 필요하다(김광섭 외, 2004). 따라서 본 연구에서는 인공위성 원격탐사 자료와 지형자료, 기상관측 자료와 같은 가용자료와 신경망(Neural Network) 모형을 이용하여 우리나라의 토양수분 분포도를 작성하고자 한다. 우선 신뢰도 높은 토양수분 관측자료를 가진 용담댐유역(6개 지점)에 대하여 전체적인 토양수분의 거동을 파악하여 토양수분 추정 모형의 적용 가능성을 확인하였다 이를 사용해서 용담댐 유역의 토양수분 분포와 우리나라 전역에 대한 토양수분 분포도를 추정하고자 한다. 신뢰할 수 있는 지상관측 토양수분 관측치가 다양한 지상조건에 대하여 존재하지 않는 한계에도 불구하고 제시된 토양수분추정 방법은 제한된 가용자료를 사용한 우리나라 전역의 토양수분 추정에 있어 합리적인 접근법이라 판단된다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2011.05a
• /
• pp.408-408
• /
• 2011

수자원관리와 수문모형에 있어 강수, 증발산, 침투, 침루 등의 물 순환과정에 대한 실질적인 이해와 분석연구의 중요도가 높아지고 있는 실정이며, 그중에서도 토양수분은 강수의 침투, 유출 등의 지표면과 대기사이의 질량 및 에너지이동에 관여하는 중요한 요소로서 수자원 및 수문현상에 직접적인 영향을 미친다. 이를 위해 강수, 증발산, 토양수분과 같은 수문변수에 대한 다양한 관측이 실시되어야 하지만 국내에서는 지속적이고 안정적으로 지상관측을 할 수 없는 실정이며 관련 기반기술도 매우 취약하다. 따라서 이를 극복하기 위해서는 위성영상자료를 이용함으로써 한반도 전체에 대한 광역적인 토양수분자료의 획득을 용이하게 한다. 본 연구의 연구유역은 수자원 연구를 위해서 지정된 용담댐 시험유역으로 하였으며, 토양수분 관측지점의 지상관측 수문자료인 각 지점별 강수량, 지면온도, 인공위성자료인 MODIS 정규식생지수 등의 가용자료를 수집하고 신경망모형을 활용한 토양수분자료 생산 모형을 개발하여, 개선된 시공간 분해능과 공간정보 대표성을 가진 광역 토양수분자료를 생산하고 적용타당성을 분석하였다. 산정된 토양수분모형의 적용가능성을 파악하고자 용담댐 유역의 각 지점별 토양수분 관측데이터와 추정데이터를 비교한 결과 추천, 부귀, 상정 지점의 경우 평균 약 0.9257의 상관계수와 약 1.2917의 평균제곱근오차를 보였고, 검증지점인 천천2의 경우 약 0.8982의 상관계수와 약 5.1361의 평균제곱근오차의 결과를 보여주었으며 토양수분 추정모형의 적용가능성이 높음을 확인할 수 있었다.

• Smart Media Journal
• /
• v.9 no.1
• /
• pp.45-50
• /
• 2020

With the development of deep learning, studies using artificial neural networks based on deep learning in recommendation systems are being actively conducted. Especially, the recommendation system based on RNN (Recurrent Neural Network) shows good performance because it considers the sequential characteristics of data. This study proposes a travel route recommendation system using GRU(Gated Recurrent Unit) and Session-based Parallel Mini-batch which are RNN-based algorithm. This study improved the recommendation performance through an ensemble of top1 and bpr(Bayesian personalized ranking) error functions. In addition, it was confirmed that the RNN-based recommendation system considering the sequential characteristics in the data makes a recommendation reflecting the meaning of the travel destination inherent in the travel route.