• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.48 no.10
• /
• pp.773-782
• /
• 2020

The development of joint FE models for deep learning neural network (DLNN)-based hybrid FEA is presented. Material models of bolts and bearings in the front axle of tractor, showing complex behavior induced by various tightening conditions, were replaced with DLNN models. Bolts are modeled as one-dimensional Timoshenko beam elements with six degrees of freedom, and bearings as three-dimensional solid elements. Stress-strain data were extracted from all elements after finite element analysis subjected to various load conditions, and DLNN for bolts and bearing were trained with Tensorflow. The DLNN-based joint models were implemented in the ABAQUS user subroutines where stresses from the next increment are updated and the algorithmic tangent stiffness matrix is calculated. Generalization of the trained DLNN in the FE model was verified by subjecting it to a new loading condition. Finally, the DLNN-based FEA for the front axle of the tractor was conducted and the feasibility was verified by comparing with results of a static structural experiment of the actual tractor.

• Journal of Korea Water Resources Association
• /
• v.42 no.7
• /
• pp.525-535
• /
• 2009

In this paper a heuristic method for identifying individual pipes in water pipe networks to determine specific sections of the pipes that need to be replaced due to deterioration. An appropriate minimum pipe length is determined by selecting the pipe length that has the greatest variance of the average cumulative break number slopes among the various pipe lengths used. As a result, the minimum pipe length for the case study water network is determined as 4 m and a total of 39 individual pipe IDs are obtained. The economically optimal replacement times of the individual pipe IDs are estimated by using the threshold break rate of an individual pipe ID and the pipe break trends models for which the General Pipe Break Prediction Model(Park and Loganathan, 2002) that can incorporate the linear, exponential, and in-between of the linear and exponetial failure trends and the ROCOFs based on the modified time scale(Park et al., 2007) are used. The maximum log-likelihoods of the log-linear ROCOF and Weibull ROCOF estimated for the break data of a pipe are compared and the ROCOF that has a greater likelihood is selected for the pipe of interest. The effects of the social costs of a pipe break on the optimal replacement time are also discussed.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.2
• /
• pp.610-619
• /
• 2016

Digital manufacturing (DM) technology helps engineers design products promptly and reliably at low production cost by simulating a manufacturing process and the material behavior of a product in use, based on three-dimensional digital modeling. The computing infrastructure for digital manufacturing, however, is usually expensive and, at present, the number of professional design engineers who can take advantage of this technology to a product design accurately is insufficient, particularly in small and medium manufacturing companies. Considering this, the Korea Institute of Science and Technology Information (KISTI) and H University is operating a DM track in the form of Industry-University-Research Institute collaboration to train high-performance-computing-based DM professionals. In this paper, a series of courses to train students to work directly into DM practice in industry after graduation is reported. The operating cases of the DM track for two years since 2013 are presented by focusing on the progress in establishment, lecture and practice contents, evaluation of students, and course quality improvement. Overall, the track management, curriculum management, learning achievement of students have been successful. By expediting more active participation of the students in the track and providing more internship and job offers in the participating companies in addition to collaborative capstone design projects, the track can be expanded by fostering a nationwide training network.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.12
• /
• pp.636-644
• /
• 2019

The purpose of this study is to explore practical intervention strategies by identifying the relationships among digital informatization capacity, level of digital informatization accessability and life satisfaction of disabled people and to determine differences among these relationships depending on perceived level of social support networks. The participants were 1,639 disabled people from the 2017 digital information gap survey and the results, based on structural equation modeling and multi-group analysis, are as follows. First, digital informatization capacity has a positive influence on the level of digital informatization accessability(β=.65), and life satisfaction(β=.08). The level of digital informatization accessability also has positive influence on life satisfaction(β=.44). Second, the analysis result of the mediated effects of digital informatization accessability level between digital informatization capacity and life satisfaction was significant at a level (β=.29) even greater than the direct effect of digital informatization capacity on life satisfaction. Third, digital information capacity and digital informatization accessability have an influence on life satisfaction regardless of their perceived level of social support. The findings suggest that creating online environments where disabled people can enjoy leisure, culture, and social interaction with high accessibility and utility are as important as providing education for improving their digital informatization capacity.

• The Journal of the Korea Contents Association
• /
• v.6 no.12
• /
• pp.1-11
• /
• 2006

Many research demonstrations and commercial applications have been tried to develop face detection and recognition systems. Human face detection plays an important role in applications such as access control and video surveillance, human computer interface, identity authentication, etc. There are some special problems such as a face connected with background, faces connected via the skin color, and a face divided into several small parts after skin region segmentation in generally. It can be allowed many face detection techniques to solve the first and second problems. However, it is not easy to detect a face divided into several parts of regions for reason of different illumination conditions in the third problem. Therefore, we propose an efficient modified skin segmentation algorithm to solve this problem because the typical region segmentation algorithm can not be used to. Our algorithm detects skin regions over the entire image, and then generates face candidate regions using our skin segmentation algorithm For each face candidate, we implement the procedure of region merging for divided regions in order to make a region using adjacency between homogeneous regions. We utilize various different searching window sizes to detect different size faces and a face detection classifier based on a back-propagation algorithm in order to verify whether the searching window contains a face or not.

• Journal of Korean Society of Transportation
• /
• v.34 no.2
• /
• pp.123-134
• /
• 2016

Traffic accident at intersections takes 44.3% of total number of accidents on entire road network of Korea in 2014. Although several studies addressed contributory factors of accidents at signalized intersection, very few is known about the factors at rural unsignalized intersections. The objective of this study is therefore to investigate specific characteristics of crashes at rural unsignalized intersection and to identify contributory factors in causing crashes by statistical approach using the Ordered Logistic Regression Model. The results show that main type of car crashes at unsignalized intersection during the daytime is T-bone crashes and the number of crashes at 4-legged intersections are 1.53 times more than that at 3-legged intersections. Most collisions are caused by negligence of drivers and violation of Right of Way. Based upon the analysis, accident severity is modeled as classified by two types such as 3-legged intersection and 4-legged intersection. It shows that contributory factors in causing crashes at rural unsignalized intersections are poor sight distance problem, average daily traffic, time of day(night, or day), angle of intersection, ratio of heavy vehicles, number of traffic violations at intersection, and number of lanes on minor street.

• Journal of Korea Water Resources Association
• /
• v.43 no.2
• /
• pp.211-220
• /
• 2010

The objective of this paper presents the application of an "integrated urban flood modeling-runoff model, urban flood model and sewer system model-" in a highly urbanized area of Samcheok where is seriously inundated in 2002 and 2003. For this, we demonstrate how couple a 1-D hydrodynamic model of the river, a 2-D hydrodynamic model of the overland (surface) flow, and a sewer network model including each boundary conditions. In order to make data file for the model, topographic information like elevation and share rate of buildings are directly extracted from DEM or topographical source data without data exchange to avoid uncertainty errors. Furthermore, the research is to assess the impacts of Manning n and buildings influences to inundated depth by changing its share ratio from 10 % to 30 % in low-land urban area. As a results, we found out that the urban inundated depth was decreased by Manning n but increased by buildings ratio. The calculated results of inundation was similar with observed one in 2002 and 2003 flooding. Furthermore, the area was also inundated under not riverbank break case in 2002 flooding.

• Journal of Internet Computing and Services
• /
• v.10 no.5
• /
• pp.65-78
• /
• 2009

The most commonly adopted approach to find valuable information from tree data is to extract frequently occurring subtree patterns from them. Because mining frequent tree patterns has a wide range of applications such as xml mining, web usage mining, bioinformatics, and network multicast routing, many algorithms have been recently proposed to find the patterns. However, existing tree mining algorithms suffer from several serious pitfalls in finding frequent tree patterns from massive tree datasets. Some of the major problems are due to (1) modeling data as hierarchical tree structure, (2) the computationally high cost of the candidate maintenance, (3) the repetitious input dataset scans, and (4) the high memory dependency. These problems stem from that most of these algorithms are based on the well-known apriori algorithm and have used anti-monotone property for candidate generation and frequency counting in their algorithms. To solve the problems, we base a pattern-growth approach rather than the apriori approach, and choose to extract maximal frequent subtree patterns instead of frequent subtree patterns. The proposed method not only gets rid of the process for infrequent subtrees pruning, but also totally eliminates the problem of generating candidate subtrees. Hence, it significantly improves the whole mining process.

• Journal of Korean Society of Environmental Engineers
• /
• v.27 no.5
• /
• pp.453-460
• /
• 2005

Since late 1970s, water quality monitoring had been performed in Korea by various agencies according to their own needs and objectives. Lacking of consistency in principles, such diverse operation has been producing difficulties in management of information on water quality parameters. These difficulties resulted from the fact that the expansion of monitoring networks had been based not on systematic methodology with considerations on scientifically established planning objectives but on personal experiences and subjective judgments in locating monitoring stations. This study aimed, therefore, to develop a selection methodology for locating monitoring stations to provide various informations on water quality to satisfy various monitoring objectives such as understanding short and long term trends of water quality, monitoring violations to water quality standards, identifying external causes and sources affecting water quality changes, supporting utilization of water resources, examining short-term variations in water quality through a concentrated investigation, estimating pollution loads from each unit watershed, supporting water quality modeling, and establishing informative systems for water resources management. Also, we applied the proposed methodology and presented an improved expansion plan for water quality monitoring networks in the Han River.

• Geophysics and Geophysical Exploration
• /
• v.12 no.4
• /
• pp.347-354
• /
• 2009

Physical properties of rocks are strongly dependant on details of pore micro-structures, which can be used for quantifying relations between physical properties of rocks through pore-scale simulation techniques. Recently, high-resolution scan techniques, such as X-ray microtomography and high performance computers make it possible to calculate permeability from pore micro-structures of rocks. We try to extend this simulation methodology to velocity and electrical conductivity. However, the smoothing effect during tomographic inversion creates artifacts in pore micro-structures and causes inaccurate property estimation. To mitigate this artifact, we tried to use sharpening filter and neural network classification techniques. Both methods gave noticeable improvement in pore structure imaging and accurate estimation of permeability and electrical conductivity, which implies that our method effectively removes the smoothing effect in pore structures. However, the calculated velocities showed only incremental improvement. By comparison between thin section images and tomogram, we found that our resolution is not high enough, and it is mainly responsible for the inaccuracy in velocity despite the successful removal of the smoothing effect. In conclusion, our methods can be very useful for pore-scale modeling, since it can create accurate pore structure without the smoothing effect. For accurate velocity estimation, the resolution of pore structure should be at least three times higher than that for permeability simulation.