• Journal of the Korean Association of Geographic Information Studies
• /
• v.6 no.1
• /
• pp.119-131
• /
• 2003

Rainfall-runoff process is under the control of hydrologic parameters having temporal and spatial variety. Accordingly, it is difficult to efficiently deal them since many parameters and various information are required to perform hydrologic simulation. So the purposes of this study is to estimate the runoff volume by frequency using GIS techniques and NRCS method. The analysis of frequency rainfall is analyzed using FARD 2002 program and the result of goodness of fit test show that Log-pearson type III is suitable distribute type for the applied area. TOPAZ program used for the analysis of DEM data examining into geological characteristic. NRCS curve numbers estimated using landuse map and soil map for the estimation of effective rain fall in the basin. The storm Type II and Type III were used as the type for the application of NRCS. The result of application show that the runoff volumes above 80 years frequency in return period have similar patterns regardless of Type II and Type III. In addition, the results of comparison with runoff volumes by frequency in the report of river improvement master plan show that it have similar volumes as the relative errors for them of 80, 100 years frequency are each 7.65%, 5.33%.

• Journal of IKEEE
• /
• v.21 no.2
• /
• pp.136-141
• /
• 2017

In this paper, we propose an algorithm for memory optimization to apply deep learning to PC. The proposed algorithm minimizes the memory and computation processing time by reducing the amount of computation processing and data required in the conventional deep learning structure in a general PC. The algorithm proposed in this paper consists of three steps: a convolution layer configuration process using a random filter with discriminating power, a data reduction process using PCA, and a CNN structure creation using SVM. The learning process is not necessary in the convolution layer construction process using the discriminating random filter, thereby shortening the learning time of the overall deep learning. PCA reduces the amount of memory and computation throughput. The creation of the CNN structure using SVM maximizes the effect of reducing the amount of memory and computational throughput required. In order to evaluate the performance of the proposed algorithm, we experimented with Yale University's Extended Yale B face database. The results show that the algorithm proposed in this paper has a similar performance recognition rate compared with the existing CNN algorithm. And it was confirmed to be excellent. Based on the algorithm proposed in this paper, it is expected that a deep learning algorithm with many data and computation processes can be implemented in a general PC.

• Geotechnical Engineering
• /
• v.1 no.1
• /
• pp.59-72
• /
• 1985

This paper deals with the numerical analysis by the (mite element method introducing Biot's theory of consolidation and the modified Cambridge model proposed by Roscoe school of Cambridge University as constitutive equation and using Christian-Boehner's technique. Especially, time interval and division of elements are investigated in vies of stability and economics. In order to check the validity of author's program, the program was tested with one-dimensional consolidation case followed by Terzaghi's exact solution and with the results of the Magnan's analysis for existing banking carried out for study at Cubzac-les-ports in France. The main conclusions obtained are summarized as follows: 1. In the case of one-dimensional consolidation, the more divided the elements are near the surface of the foundation, the higher the accuracy of the numerical analysis is. 2. For the time interval, it is stable to divide 20 times per 1-lg cycle. 3. At the element which has long drain distance, the Mandel-fryer effect appears due to time lag. 4. Lateral displacement at an initial loading stage predicted by author's program, in which the load was assumed as not concentrative. but rather in grid form, is well consistent with the value of observation. 5. The pore water pressure predicted by author's program has a better accordance with the value of observation compared with Magnan's results. 6. Optimum construction control by Matsuo-Kawamura's method is possible with the predicted lateral displacement and settlement by the program.

• Environmental Mutagens and Carcinogens
• /
• v.25 no.2
• /
• pp.51-59
• /
• 2005

Overexpression of HSP25 delayed cell growth, increased the level of $p21^{waf}$, reduced the levels of cyclin D1, cylcin A and cdc2, and induced radioresistance in L929 cells. We demonstrated that extracellular regulated kinase (ERK) and MAP kinase/ERK kinase (MEK) expressions as well as their activation (phospho-forms) were inhibited by hsp25 overexpression. To confirm the relationship between ERK1/2 and hsp25-mediated radioresistance, ERK1 or ERK2 cDNA was transiently transfected into the hsp25 overexpressed cells and their radioresistance was examined. HSP25-mediated radioresistance was abolished by overexpression of ERK2, but not by overexpression of ERK1. Alteration of cell cycle distribution and cell cycle related protein expressions (cyclin D, cyclin A and cdc2) by hsp25 overexpression were also recovered by ERK2 cDNA transfection. Increase in Bc1-2 protein by hsp25 gene transfection was also reduced by subsequent ERK2 cDNA-transfection. In addition, HSP25 overexpression reduced reactive oxygen species (ROS) and increased expression of manganese superoxide dismutase (MnSOD) gene. Increased activation of NF-kB (IkB degradation) was also found in hsp25-overexpressed cells. Moreover, transfection of hsp25 antisense gene abrogated all the HSP25-mediated phenomena. To further elucidate the exact relationship between MnSOD induction and NF-kB activation, dominant negative $I-kB\alpha(I-kB\alpha-DN)$ construction was transfected to HSP25 overexpressed cells. $I-kB\alpha-DN$ inhibited HSP25 mediated MnSOD gene expression. In addition, HSP25 mediated radioresistance was blocked by $I-kB\alpha-DN$ transfection. Blockage of MnSOD with antisense oligonucleotides in HSP25 overexpressed cells, prevented apoptosis and returned the ERK1/2 activation to the control level. From the above results, we suggest for the first time that reduced oxidative damage by HSP25 was due to MnSOD-mediated down regulation of ERK1/2.

• Journal of the Korea Concrete Institute
• /
• v.21 no.4
• /
• pp.465-471
• /
• 2009

To prevent the explosive spalling of the high strength concrete and control the rise of temperature in the steel rebar during fire, a fiber cocktail method has been proposed simultaneously with the use of polypropylene and steel fiber. After applying the fiber cocktail (polypropylene and steel fibers) into the mixture of high strength concrete with a compressive strength of between 40 and 100 MPa and evaluating the thermal properties at elevated temperatures, the fire test was carried out on structural members in order to evaluate the fire resistance performance. Two column specimens were exposed to the fire without loading for 180 minutes based on the standard curve of ISO-834. No explosive spalling has been observed and the original color of specimen surface was changed to light pinkish grey. The inner temperature of concrete dropped rapidly starting from 60mm deep. After 60 minutes of exposure to the fire, the temperature gradient of fiber cocktail reinforced high strength concrete was measured as 2.2oC/mm, which is approximately 5 times less than that of normal concrete. The final temperatures of steel rebar after 180 minutes of fire test resulted in 488.0oC for corner rebar, 350.9oC for center rebar, and 419.5oC for total mean of steel rebar. The difference of mean temperature between corner and center rebar was 137.1oC The tendency of temperature rise in concrete and steel rebar changed between 100oC and 150oC The cause of decrease in temperature rise was due to the water vaporization in concrete, the lower temperature gradient of the concrete with steel and polypropylene fiber cocktails, the moisture movement toward steel rebars and the moisture clogging.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2008.11a
• /
• pp.777-780
• /
• 2008

According to the recent tendency that the buildings in the downtown are gradually Manhattanized, the very thick massive concrete is selected as the foundation of architectures. By the way, because this mass concrete cannot be simultaneously pour in a great quantity due to the circumstance at the field, not only the questions on the unification between the concretes pour on the upper layer and the lower layer are presented but also the cracks by the internal force from the difference of hydration exothermic period are occurred, which are pointed out as the problems. Thus, this study performed Mock-up test to apply the hydration heat controlling method of massive concrete for horizontal partition pouring construction to the building sites for the purpose of securing the stability on the cracks by the internal force from the difference of hydration exothermic period on the upper layer and the lower layer of massive concrete and checked the efficiency. As the results of test, in case of setting time difference method by super retarder with 2 layers and 4 layers, the effect that temperature gaps between upper part and lower part were lowered and the possibility of crack occurrence was decreased as the peak time of the heat of hydration became delayed to the latter term could be confirmed.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.20 no.2
• /
• pp.17-31
• /
• 2017

The purpose of this study was to carry out a comprehensive assessment of Donjibong Park in Daegu and establish a master plan for green space considering the functional characteristics of the park. For this purpose, green space in the park was assessed based on 13 indicators using three functions. A comprehensive green master plan was then proposed. Green spaces with high scores for the natural-ecological and environmental-control functions were found in the central area of the park, whereas green spaces with low scores were found near the park boundary and paved areas. Green spaces in the southeast were found to have a high score for usage functions due to their proximity to a residential area. In the master plan based on the comprehensive assessment, the plan space was divided into a conservation area, restoration area, buffer area, environmental monitoring area, observatory/trail/facility area, and waterside green space area. Also, concrete plan directions were proposed based on each plan space. These results can be utilized for establishing objective plans in the future construction of new parks.

• The KIPS Transactions:PartD
• /
• v.11D no.4
• /
• pp.937-950
• /
• 2004

With growing population of internet and web applications, distributed environment is considered to be the standard architecture of application. A network management systems(NMS) is necessary to control and monitor the complex network resources for providing and sharing the heft quality service. We recognize the NMS as a standard infrastructure for supporting efficient networking and a separate commercial applications. We believe every resource including software, hardware and environment for the network management should be separated from special protocols, vendors and applications. Therefore, We need a standard network management system that is efficient and consistent because of the heterogeous network features. In regards to software development, software reuse through assembling and extending the reusable elements such as patterns and components assures to realize the best productivity and quality The component based development(CBD) methodology that can assemble black box though well defined interfaces makes it possible to develop easer and quicker applications and is proved as the best software development solution involved in construction, selection and assembly of components. In this thesis, we describe the architecture for the network management and identify, define and design the components through analysis and design in the network management domain and Identified components mapped to the component architecture. We also specify the component development and design and implement the component for developing the network management. Implemented components apply to the component repository system that register, retrieve and understand the components. We analyze, design and implement the entire network management system based on configuration, connection, performance and fault management through the pre-developed components.

• Journal of the Korea Concrete Institute
• /
• v.17 no.3 s.87
• /
• pp.455-464
• /
• 2005

Today the application of antiwashout underwater concrete to the construction sites is increasing steadily, while its reliability is in issue. Particularly, antiwashout underwater concrete is known to have very weak durability on frost attack, and hence Japan society of civil engineers(JSCE) regulated that not to use of antiwashout underwater concrete where the freezing and thawing is suspected. This study aims the improvement of the freezing and thawing resistance for antiwashout underwater concrete. From the results of fundamental test, FA20 and SG50 showed good performance in fluidity and long term compressive strength than control concrete. Meanwhile, MK10 marked the highest compressive strength through the whole curing age but a defect on fluidity was discovered. The results from the repeated freezing and thawing test show that the large volumes of air entrapped by cellulose based antiwashout underwater admixture gave bad effects to frost durability and hence not much benefits were confirmed from the use of mineral admixtures. However there were some increasing effects on frost durability of MK10 and SG50 by securing $6{\pm}0.5\%$ of entraining air. In the meantime, there was a increasing tendency of frost durability by increasing blame's fineness of ground granulated blast furnace slag.

• Journal of the Korea Concrete Institute
• /
• v.19 no.4
• /
• pp.467-474
• /
• 2007

In this paper, a total of 13 beams with bonding, anchorage system, amount of prestressing and span length as variables of experiment were tested in flexural test and analyzed in finite element analysis; one control beam, two simplified FRP-boned beams, four prestressed FRP-unbonded beams and four prestressed FRP-bonded beams. Also, a nonlinear finite element analysis of beams in the flexural test is performed by DIANA program considered material nonlinear of concrete, reinforcement and the interfacial bond-slip model between concrete and CFRP plates. The failure mode of prestressed CFRP plated-beams is not debonding but FRP rupture. RC members strengthened with external bonded prestressed CFRP plates occurred 1st and 2nd debonding of the composite material. After the debonding of CFRP plates occurs in bonded system, behavior of bonded CFRP-plated beams change into that of unbonded CFRP-plated beams due to fix of the anchorage system. Also, It was compared flexural test results and analytical results of RC members strengthened with CFRF plates. The ductility of beams strengthened by CFRP plates with the anchorage system is considered high with the ductility index of above 3. Analysis results showed a good agreement with experiment results in the debonding load, yield load and ultimate load.