• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.10 no.3
• /
• pp.14-24
• /
• 2007

The purpose of this study was to investigate how the landscape characteristics and the physical factors of landscape would affect the preference for the Gapcheon in Daejeon Metropolitan City. The Gapcheon was divided in three sections of the outskirts, Expopark areas, and residential complexes. After selecting seven landscape points where the sections could be expressed best, photographs were taken both in the upstream and downstream direction. The questionnaire used to evaluate the river's landscape included 20 items of adverbs that described the form of the river and one item to rate the overall preference. By analyzing the 14 pictures taken, the occupancy rates of the landscape elements in terms of the sky, river, vegetation of the river, mountain, and artificial structures. Image factor analysis was conducted for each of the sections in order to analyze the landscape characteristics of the Gapcheon, and then regression analysis was conducted in order to analyze the relationships among the physical factors influencing the preference of the landscapes. The results were as follows : Factors that compose the visual characters of urban river were classified be the aesthetic factor, the emotional factor and the situation factor. These 3 factors showed a 65.8% total variance. The river landscape with the biggest preference was the one from the Daedeok Grand Bridge as the occupancy area of the mountain, sky, and river was large and distributed evenly and the vegetation of the river was in a good harmony with the surroundings. After carrying out regression analysis to examine the relationships between the visual preference of Gapcheon and the physical factors of landscape(the sky, river, vegetation of the river, mountain, and artificial structure), the following regressions model was made : PRE=5.906+0.017(river)-0.053(artificial structure)-0.060(vegetation of the river) (R-square=0.48).

• Journal of Korean Society on Water Environment
• /
• v.23 no.5
• /
• pp.676-682
• /
• 2007

The bacterial number, extracellular enzyme activities and structure of bacterial community which are major constituent of aquatic ecosystem within the artificial vegetation island (AVI) were compared to those of the nearby pelagic lake waters in order to evaluate the possibility of the AVI as a eco-technological measure for water quality improvement and restoration of littoral zone in man-made reservoirs. There was not a significant difference in the total number of bacteria, but the number of active (viable) bacteria within the AVI was about 0.7 to 4.1 times higher than nearby pelagic lake water. The ratio of the number of active bacteria versus the total number of bacteria was also higher in the AVI than nearby pelagic lake water. The activities of ${\beta}$-glucosidase and phosphatase were 1.0 to 13.1 and 0.8 to 7.3 times higher respectively in the AVI than nearby pelagic lake water, showing that microorganisms were more active within the AVI. The bacterial communities of the two waters, examined by FISH method, did not indicate a clear difference in the springtime when the growth of macrophytes was immature, but during summer and fall it showed a clear difference indicating the formation of distinct bacterial community within the AVI compared to nearby lake water. From the results of this study, we conclude that AVI can contribute to make up the littoral ecosystem which show rapid cycling of matters through active detritus food chain in the dam reservoirs which have unstable aquatic ecosystem due to short hydraulic residence time and to strengthen the self-purification capacity of the lake.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.25 no.11
• /
• pp.787-797
• /
• 2015

In this paper, a cochlea-inspired artificial filter bank(CAFB) was developed in order to efficiently acquire dynamic response of structure, and it was also evaluated via dynamic response experiments. To sort out signals containing significant modal information from all the dynamic responses of structure, it was made to adopt a band-pass filter optimizing algorithm(BOA) and a peak-picking algorithm (PPA). Optimally designed on the basis of El-centro and Kobe earthquake signals, it was then embedded into the wireless multi-measurement system(WiMMS). In order to evaluate the performance of the developed CAFB, a vibration test was conducted using the El-centro and Kobe earthquake signals, and structural responses of a two-span bridge were obtained and analyzed simultaneously by both a wired measurement system and a CAFB-based WiMMS. The test results showed that the compressed dynamic responses acquired by the CAFB-based WiMMS matched with those of the wired system, and they included significant modal information of the two-span bridge. Therefore this study showed that the developed CAFB could be used as a new, economic, and efficient measurement device for wireless sensor networks(WSNs) based real time structural health monitoring because it could reconstruct the whole dynamic response using the compressed dynamic response with significant modal information.

• Korean Journal of Environment and Ecology
• /
• v.12 no.4
• /
• pp.331-344
• /
• 1999

To study plant community structure and successional sere of the Kojedo District, Hallyo-Haesang National Park, fifty six plots were set up and surveyed(each plot $l00\m^2$). Fifty-six plots were classified into four community, Quercus variabilis-Zelkova serrata community(I), Q. variabilis-Cinnamomum japonicum-Neolitsea sericea community(II), Pinus thunbergii-C. japonicum-N. sericea community(III), P. thunbergii-P. densiflora community(IV) by TWINSPAN technique. Most area of forests in Kojedo district are artificial or coppice forest after artificial disturbance. The age of four community were about 35~40 years and the number of species 32~44 species per $600\m^2$. Ecological succession had been proceeded from deciduous broad-leaved species(Q. variabilis-Z. serrata etc.), P. thunbergii, P. densiflora to evergreen broad leaved forest(C. japonicum, N. sericea etc.).

• Journal of the Korean Geotechnical Society
• /
• v.36 no.8
• /
• pp.7-15
• /
• 2020

Constructing the maximum displacement and shear force database for the seismic performance of building with soil-structure interaction under varied earthquake scenarios and geotechnical conditions is critical in developing the neural network-based prediction models. However, using the available 3D FEM-based computer simulation techniques causes high computation costs in developing the database. This study introduces the framework of developing the artificial neural network (ANN) model to predict the seismic performance of building at given Poisson's ratio and shear wave velocity of soil. The simple Single-Degree-Of-Freedom system was used to develop the database and the performance of the developed neural network model is discussed through the evaluated coefficient of determination (R²). In addition, ANN models were developed for 90~100% percentile of the database to assess the accuracy of the developed ANN models in each percentile.

• Journal of Biomedical Engineering Research
• /
• v.24 no.4
• /
• pp.301-307
• /
• 2003

In this study, we Presented computational results on the blood flow in the sac of the Korean artificial heart. Two dimensional unsteady flow was assumed and we utilized a finite element commercial code ADINA to simulate the blood flow. Rigid body-solid contact were considered between the actuator and the blood sac and fluid-structure interaction between the blood and the sac. The three geometric models proposed in the design process were simulated to assess the hemodynamic characteristics of the models According to the computational results, a strong flow to the outlet and a stagnated flow region near the inlet were observed during systole. The sac was filled with blood and recirculating flow was generated near the outlet during diastole. Shear stress during systole had its extreme values near the outlet edge whereas the magnitude of shear stress values were relative)v high near the inlet edge and the contacting surface with the actuator.

• Journal of The Geomorphological Association of Korea
• /
• v.19 no.2
• /
• pp.69-79
• /
• 2012

The aim of this study is to analyze the mid-term changes of beach and dune morphology at Sohwang beach, Korea using by Total Station. Measurements executed 4 times during two year. Based on ArcInfo as point cloud obtained through precise measurement data by Total Station, alteration of beach and dune was analyzed at DEM, of which cell size is about 1m. Since these artificial constructions have influenced current systems of this region, the large-scale sand movements above mentioned have occurred around the jetty and the sea-wall. There occurred sedimentation in the north of the Jetty and erosion in the south of the Jetty, which is installed at the central part of object area. The direction of recent topographic development does not coincide with that of wind, and, rather, topographic changes occurred mainly at beaches and dunes due to the transformation of coastal water flow caused by artificial structure nearby. If precise measurement is conducted periodically, and long term monitoring is carried out by installing equipment measuring movement pattern of sediment around artificial structure, cause of topographic change around the object area could be discovered.

• Journal of Digital Convergence
• /
• v.19 no.10
• /
• pp.253-263
• /
• 2021

Machine learning is the process of constructing a cost function using learning data used for learning and an artificial neural network to predict the data, and finding parameters that minimize the cost function. Parameters are changed by using the gradient-based method of the cost function. The more complex the digital signal and the more complex the problem to be learned, the more complex and deeper the structure of the artificial neural network. Such a complex and deep neural network structure can cause over-fitting problems. In order to avoid over-fitting, a weight decay regularization method of parameters is used. We additionally use the value of the cost function in this method. In this way, the accuracy of machine learning is improved, and the superiority is confirmed through numerical experiments. These results derive accurate values for a wide range of artificial intelligence data through machine learning.

• Journal of Korea Technology Innovation Society
• /
• v.22 no.3
• /
• pp.340-361
• /
• 2019

This study analyzes the technological structure of artificial intelligence (AI) and technological capabilities of AI companies based on patent information. 2589 AI patents registered in USPTO from 2007 to 2017 were collected and analyzed by the Latent Dirichlet Allocation (LDA) to derive 20 AI technology topics. Analysis of technology development trends by AI technology reveals that visual understanding, data analysis, motion control, and machine learning are growing, while language understanding and speech technology are sluggish. In addition, we also investigated leading companies in each sub-field of AI as well as core competencies of global IT companies. The findings of this study are expected to be fruitfully used for formulation and implementation of technology strategy of AI companies.

• Current Applied Physics
• /
• v.18 no.11
• /
• pp.1320-1326
• /
• 2018

High crystalline quality CrN thin films have been grown on $La_{0.67}Sr_{0.33}MnO_3$ (LSMO) templates by molecular beam epitaxy. The structure and magnetic properties of CrN/LSMO heterojunctions are investigated combining with the experiments and the first-principles simulation. The N?el temperature of the CrN/LSMO samples is found to be 281 K and the saturation magnetization of CrN/LSMO increases compared to that of LSMO templates. The magnetic property of CrN/LSMO heterostructures mainly comes from Cr atoms of (001) CrN and Mn atoms of (001) LSMO. The (001) LSMO induces and couples the spin of the CrN sublattice at CrN/LSMO interface.