• Geomechanics and Engineering
• /
• 제12권4호
• /
• pp.579-593
• /
• 2017

Geosynthetic reinforced soil retaining walls can be employed as railway embankments to carry large static and dynamic train loads, but very few studies can be found in the literature that investigate their dynamic behavior under simulated wheel loading. A large-scale dynamic test on a reinforced soil railway embankment was therefore carried out. The model embankment was 1.65 meter high and designed to have a soilbag facing. It was reinforced with HDPE geogrid layers at a vertical spacing of 0.3 m and a length of 2 m. The dynamic test consisted of 1.2 million cycles of harmonic dynamic loading with three different load levels and four different exciting frequencies. Before the dynamic loading test, a static test was also carried out to understand the general behavior of the embankment behavior. The study indicated the importance of loading frequency on the dynamic response of reinforced soil railway embankment. It also showed that toe resistance played a significant role in the dynamic behavior of the embankment. Some limitations of the test were also discussed.

• 대한전기학회논문지
• /
• 제35권8호
• /
• pp.333-340
• /
• 1986

This paper presents an approach for determining the discrete reduced-order models for largescale system by using matrix sign function. We define projection operators based on the matrix sign function and develop the algorithm for model-reduction by using them. Simulation studies show that the proposed altgorithm is very useful.

• 농촌계획
• /
• 제7권1호
• /
• pp.61-76
• /
• 2001

During the performance of large scale tidal land reclamation project along the coast line, the construction of large scale structures such as sea-dikes, closing estuaries will induce big changes of near-shore hydraulic behavior. In this paper, its is aimed to verify the change of tide and currents after the construction of sea-dike of the Saemangeum project along the coast line. Numerical scale model "TRISULA" which development by Delft Hydraulics in the Netherlands was used. "TRISULA" is adopting the finite difference numerical scheme, and mostly using for hydro-dynamic solution along the sea and estuaries. Model boundary is covering $100{\times}170$ Km and constructed with $133{\times}337$ grids. Outer side boundary is divided 48 sections, and input 37 tidal components are gained from another big scale numerical "Yellow Sea" model. Model calibrations & verifications were performed th field tide & current datas which were measured along sea-dike alignment during Aug. $1997{\sim}Apr$. 1999. And then, numerical simulation with the tide condition dated 17 Apr. 1999 was performed with & without sea-dike construction condition for the comparison. Evaluated boundary is 20 km out-side from sea-dike alignment. Four cross lanes were set up, each of lane contains 3 points, for the comparison of sea-dike construction effects. Results showing the tidal amplitude is reducing approximately 20 cm after the construction of sea-dike during spring tide condition, amplitude 6.9m. Currents after construction of sea-dikes along the alignment, the northern part shows 50%(inner), 90%(outer) and the southern part shows 10%(inner) 50%(outer) of the currents before construction.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권3호
• /
• pp.771-791
• /
• 2022

Protein-protein interaction (PPI) extraction from original text is important for revealing the molecular mechanism of biological processes. With the rapid growth of biomedical literature, manually extracting PPI has become more time-consuming and laborious. Therefore, the automatic PPI extraction from the raw literature through natural language processing technology has attracted the attention of the majority of researchers. We propose a PPI extraction model based on the large pre-trained language model and adversarial training. It enhances the learning of semantic and syntactic features using BioBERT pre-trained weights, which are built on large-scale domain corpora, and adversarial perturbations are applied to the embedding layer to improve the robustness of the model. Experimental results showed that the proposed model achieved the highest F1 scores (83.93% and 90.31%) on two corpora with large sample sizes, namely, AIMed and BioInfer, respectively, compared with the previous method. It also achieved comparable performance on three corpora with small sample sizes, namely, HPRD50, IEPA, and LLL.

• Fisheries and Aquatic Sciences
• /
• 제10권3호
• /
• pp.159-169
• /
• 2007

The complex 2-dimensional movements of fish during an annual migration circuit were generated and simulated by a chaotic model of fish movement, which was expanded from a small-scale movement model. Fish migration was modeled as a neural network including stimuli, central decision-making, and output responses as variables. The input stimuli included physical stimuli (temperature, salinity, turbidity, flow), biotic factors (prey, predators, life cycle) and landmarks or navigational aids (sun, moon, weather), values of which were all normalized as ratios. By varying the amplitude and period coefficients of the klinokinesis index using chaotic equations, model results (i.e., spatial orientation patterns of migration through time) were represented as fish feeding, spawning, overwintering, and sheltering. Simulations using this model generated 2-dimesional annual movements of sea bream migration in the southern and western seas of the Korean Peninsula. This model of object-oriented and large-scale fish migration produced complicated and sensitive migratory movements by varying both the klinokinesis coefficients (e.g., the amplitude and period of the physiological month) and the angular variables within chaotic equations.

• Water Engineering Research
• /
• 제5권2호
• /
• pp.81-99
• /
• 2004

Long term streamflow regime under virtual climate change scenario was examined. Rainfall forecast simulation of the Canadian Global Coupled Model (CGCM2) of the Canadian Climate Center for modeling and analysis for the IPCC SRES B2 scenario was used for analysis. The B2 scenario envisions slower population growth (10.4 billion by 2010) with a more rapidly evolving economy and more emphasis on environmental protection. The relatively large scale of GCM hinders the accurate computation of the important streamflow characteristics such as the peak flow rate and lag time, etc. The GCM rainfall with more than 100km scale was downscaled to 2km-scale using the space-time stochastic random cascade model. The HEC-HMS was used for distributed hydrologic model which can take the grid rainfall as input data. The result illustrates that the annual variation of the total runoff and the peak flow can be much greater than rainfall variation, which means actual impact of rainfall variation for the available water resources can be much greater than the extent of the rainfall variation.

• KIEAE Journal
• /
• 제4권3호
• /
• pp.27-35
• /
• 2004

The conventional way to acquire sufficient amount of daylight in interiors is to provide large openings with clear glass. The use of clear glass on the whole facade, however, might cause a sort of visual problem because of the harness of direct sun and brighter sky surface than expected. They should be filtered in opticalway or bounced in the architectural. One of the common solutions for the problem might be the use of photometric glasses with various transmittances for the glass walls. This paper deals with performance data related to the impact of various transmittal glazing materials for window systems in terms of daylighting. A series of scale model measurements was carried out with the fundamental configuration of a commonly used all-glass facades. Additionally some experimental performance index was issued for the better expression of the need of natural lighting.

• 한국원자력학회:학술대회논문집
• /
• 한국원자력학회 2001년도 춘계학술발표회요약집
• /
• pp.503.2-503
• /
• 2001

• 대한기계학회논문집B
• /
• 제30권11호
• /
• pp.1101-1106
• /
• 2006

To investigate a relation between fine scale eddies and particle dispersion in a near-wall turbulence, direct numerical simulations of turbulent channel flow laden particle are performed for $Re_{\tau}$=180. The motions of 0,8 million particles are calculated for several particle response times ($t_p$) which is the particle response time based on stokes’ friction law. The number density of particles has a tendency to increase with approaching the near-wall regions ($y^+$<20) except for cases of very small and large particle response times (i.e. $t_p$=0.02 and 15). Near the wall, the behavior and distribution of particles are deeply associated with the fine scale eddies, and are dependent on particle response times and a distance from the wall. The Stokes number that causes preferential distribution in turbulence is changed by a distance from the wall. The influential Stokes number based on the Burgers' vortex model is derived by using the time scale of the fine scale eddies. The influential Stokes number is also dependent on a distance from the wall and shows large value in the buffer layer.

• Journal of Multimedia Information System
• /
• 제5권2호
• /
• pp.111-114
• /
• 2018

Traffic accident analysis is important to reduce the occurrence of the accidents. In this paper, we analyze the traffic accident with Apriori algorithm to find out an association rule of traffic accident in Korea. We first design the traffic accident analysis model, and then collect the traffic accidents data. We preprocessed the collected data and derived some new variables and attributes for analyzing. Next, we analyze based on statistical method and Apriori algorithm. The result shows that many large-scale accident has occurred by vans in daytime. Medium-scale accident has occurred more in day than nighttime, and by cars more than vans. Small-scale accident has occurred more in night time than day time, however, the numbers were similar. Also, car-human accident is more occurred than car-car accident in small-scale accident.