• Journal of Korean Tunnelling and Underground Space Association
• /
• v.21 no.3
• /
• pp.419-432
• /
• 2019

Most of deep learning model training was proceeded by supervised learning, which is to train labeling data composed by inputs and corresponding outputs. Labeling data was directly generated manually, so labeling accuracy of data is relatively high. However, it requires heavy efforts in securing data because of cost and time. Additionally, the main goal of supervised learning is to improve detection performance for 'True Positive' data but not to reduce occurrence of 'False Positive' data. In this paper, the occurrence of unpredictable 'False Positive' appears by trained modes with labeling data and 'True Positive' data in monitoring of deep learning-based CCTV accident detection system, which is under operation at a tunnel monitoring center. Those types of 'False Positive' to 'fire' or 'person' objects were frequently taking place for lights of working vehicle, reflecting sunlight at tunnel entrance, long black feature which occurs to the part of lane or car, etc. To solve this problem, a deep learning model was developed by simultaneously training the 'False Positive' data generated in the field and the labeling data. As a result, in comparison with the model that was trained only by the existing labeling data, the re-inference performance with respect to the labeling data was improved. In addition, re-inference of the 'False Positive' data shows that the number of 'False Positive' for the persons were more reduced in case of training model including many 'False Positive' data. By training of the 'False Positive' data, the capability of field application of the deep learning model was improved automatically.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.9
• /
• pp.1172-1179
• /
• 2020

With the establishment of the railway integrated radio network (LTE-R) environment, radio-based train control transmission and reception and various forms of service are provided. The smooth delivery of these services requires improved performance in a highly reliable and available wireless environment. This paper measured the LTE-R radio communication environment to improve radio communication performance of railway integrated wireless network reliability and availability, analyzed the results, and established the wireless environment model. Based on the built-up model, we also proposed an improved radio-access algorithm to control trains for improved reliability, suggesting a way to improve stability for handover that occur during open-air operation, and proposed an algorithm for frequency auto-heating to improve availability. For simulation, data were collected from the Korea Rail Network Authority (Daejeon), Manjong-Gangneung KTX route, which can measure the actual data of LTE-R wireless environment, and the results of the simulation show performance improvement through algorithm.

• Journal of the Korea Society for Simulation
• /
• v.31 no.1
• /
• pp.11-18
• /
• 2022

Recently, research to classify human activity using imagery has been actively conducted for the purpose of crime prevention and facility safety in public places and facilities. In order to improve the performance of human activity classification, most studies have applied deep learning based-transfer learning. However, despite the increase in the number of backbone network models that are the basis of deep learning as well as the diversification of architectures, research on finding a backbone network model suitable for the purpose of operation is insufficient due to the atmosphere of using a certain model. Thus, this study applies the transfer learning into recently developed deep learning backborn network models to build an intelligent system that classifies human activity using imagery. For this, 12 types of active and high-contact human activities based on sports, not basic human behaviors, were determined and 7,200 images were collected. After 20 epochs of transfer learning were equally applied to five backbone network models, we quantitatively analyzed them to find the best backbone network model for human activity classification in terms of learning process and resultant performance. As a result, XceptionNet model demonstrated 0.99 and 0.91 in training and validation accuracy, 0.96 and 0.91 in Top 2 accuracy and average precision, 1,566 sec in train process time and 260.4MB in model memory size. It was confirmed that the performance of XceptionNet was higher than that of other models.

• Korean Journal of Construction Engineering and Management
• /
• v.23 no.1
• /
• pp.46-53
• /
• 2022

The railway platform is an important facility closely related to the safety of passengers, trains, and images of railway facilities, and requires thorough facility management. However, the problem that the finishing material (plastering mortar) for the joint finishing of dissimilar materials (concrete+granite) falls off in the direction of the track at the platform podium is occurring multiple times across the country. Since these problems threaten the safety of train operation and the safety of passengers, immediate and continuous management is required. This study tried to derive improvement plans through the analysis of the drop-off problem of finishing materials occurring at the platform podium. The status of missing finishing materials for the platform podiums of about 200 railway stations and the related design and construction standards of the Korea National Railway were investigated. After that, the cause of the drop-off of the finishing material was analyzed, and as a result, it was found that the main cause was the boundary between the roadbed and the architectural process that occurred during construction. Subsequently, in connection with the derived causes and design, construction standards, (1) improvement of finishing materials or construction methods, (2) design of finishing materials that are easy to adjust height, (3) design of separate finishing methods, (4) improvement methods and durability were suggested.

• Smart Structures and Systems
• /
• v.29 no.4
• /
• pp.625-640
• /
• 2022

Maglev rail joints are vital components serving as connections between the adjacent F-type rail sections in maglev guideway. Damage to maglev rail joints such as bolt looseness may result in rough suspension gap fluctuation, failure of suspension control, and even sudden clash between the electromagnets and F-type rail. The condition monitoring of maglev rail joints is therefore highly desirable to maintain safe operation of maglev. In this connection, an online damage detection approach based on three-dimensional (3D) convolutional neural network (CNN) and time-frequency characterization is developed for simultaneous detection of multiple damage of maglev rail joints in this paper. The training and testing data used for condition evaluation of maglev rail joints consist of two months of acceleration recordings, which were acquired in-situ from different rail joints by an integrated online monitoring system during a maglev train running on a test line. Short-time Fourier transform (STFT) method is applied to transform the raw monitoring data into time-frequency spectrograms (TFS). Three CNN architectures, i.e., small-sized CNN (S-CNN), middle-sized CNN (M-CNN), and large-sized CNN (L-CNN), are configured for trial calculation and the M-CNN model with excellent prediction accuracy and high computational efficiency is finally optioned for multiple damage detection of maglev rail joints. Results show that the rail joints in three different conditions (bolt-looseness-caused rail step, misalignment-caused lateral dislocation, and normal condition) are successfully identified by the proposed approach, even when using data collected from rail joints from which no data were used in the CNN training. The capability of the proposed method is further examined by using the data collected after the loosed bolts have been replaced. In addition, by comparison with the results of CNN using frequency spectrum and traditional neural network using TFS, the proposed TFS-CNN framework is proven more accurate and robust for multiple damage detection of maglev rail joints.

• Journal of the Korea Society for Simulation
• /
• v.18 no.3
• /
• pp.23-34
• /
• 2009

In the railroad rolling stock depot, long-term maintenance tasks is done regularly every two or four year basis to maintain the functionality of equipments and rolling stock body or for the repair operation of the heavily damaged rolling stocks by fatal accidents. This paper addresses the computer simulation model building for the rolling stock maintenance shop for the CDC(Commuter Diesel Car) and Generator Car planned to be constructed at Daejon Rolling Stock Depot, which will be moved from Yongsan Rolling Stock Depot. We evaluated the processing capacity of two layout design alternatives based on the maintenance process chart through the developed simulation models. The performance measures are the number of processed cars per year, the cycle time, shop utilization, work in process and the average number waiting car for input. The simulation result shows that one design alternative outperforms another design alternative in every aspect and superior design alternative can process total 340 number of trains per year 15% more than the proposed target within the current average cycle time.

• Journal of the Korea Safety Management & Science
• /
• v.25 no.2
• /
• pp.71-83
• /
• 2023

Since last year, the government has enforced the 'Act on the Punishment of Severe Accidents, Etc.' (hereafter referred to as the 'Serious Accident Punishment Act'), which punishes business owners and business managers who fail to fulfill their duty of safety measures with 'imprisonment of one year or more' and the Occupational Safety and Health Act. Based on this, various occupational safety and health policies were developed, including the operation of a system related to entrusting the work of safety managers. Despite these efforts, the effect of implementing the Severe Accident Punishment Act is a groundbreaking change in the current disaster prevention policy, which has increased by 0.02%P and 0.03‱P, respectively, from the previous year to 0.65% of the total accident rate and 1.10‱ of the death rate per 10,000 people as of 2022. As the need emerged, attention was paid to 'collaboration and governance with safety management institutions' in the 'Severe Disaster Reduction Roadmap' announced by the Ministry of Employment and Labor in November 2022. In this study, a meaningful result was derived by comparing and analyzing the industrial accident status of workplaces entrusted by "A" safety management institutions with the national average based on the industrial accident survey table, and the types of industrial accidents that occurred in consigned workplaces were selected as intensive management targets. The policy direction for industrial accident prevention was established. It is necessary to develop safety management work manuals based on the results of this study, expertise, discover best cases of risk assessment and develop guides, and educate and train consigned workers. In addition, it suggests that the government's guidance and supervision are needed to advance the professionalism of safety management entrusted tasks, and that safety management institutions should strengthen their roles and functions for preventing and reducing industrial accidents. However, due to difficulties in disclosing information of specialized safety management institutions, the limitation of the provision, collection, and viewing of research-related data to "A" specialized safety management institutions remains a limitation of the research. It seems likely that more thorough research will be conducted.

• Journal of The Korean Astronomical Society
• /
• v.56 no.2
• /
• pp.169-185
• /
• 2023

The GMT-Consortium Large Earth Finder (G-CLEF) is the first instrument for the Giant Magellan Telescope (GMT). G-CLEF is a fiber feed, optical band echelle spectrograph that is capable of extremely precise radial velocity measurement. G-CLEF Flexure Control Camera (FCC) is included as a part in G-CLEF Front End Assembly (GCFEA), which monitors the field images focused on a fiber mirror to control the flexure and the focus errors within GCFEA. FCC consists of an optical bench on which five optical components are installed. The order of the optical train is: a collimator, neutral density filters, a focus analyzer, a reimager and a detector (Andor iKon-L 936 CCD camera). The collimator consists of a triplet lens and receives the beam reflected by a fiber mirror. The neutral density filters make it possible a broad range star brightness as a target or a guide. The focus analyzer is used to measure a focus offset. The reimager focuses the beam from the collimator onto the CCD detector focal plane. The detector module includes a linear translator and a field de-rotator. We performed thermoelastic stress analysis for lenses and their mounts to confirm the physical safety of the lens materials. We also conducted the global structure analysis for various gravitational orientations to verify the image stability requirement during the operation of the telescope and the instrument. In this article, we present the opto-mechanical detailed design of G-CLEF FCC and describe the consequence of the numerical finite element analyses for the design.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.28 no.2
• /
• pp.61-68
• /
• 2024

The majority of high-speed railway bridges along the domestic Gyeongbu and Honam lines feature a PSC-box type structure with a span length ranging from 35 to 40m, which typically exhibits a first bending natural frequency of approximately 4 to 5Hz. When KTX high-speed trains transverse these bridges at speeds ranging from 290 to 310km/h, the vibration induced by the trains approaches the first bending natural frequency of the bridge. Furthermore, with the upcoming operation of a EMU-320 high-speed train and the anticipated increase in the speeds of these high-speed trains, there is a need to analyze the dynamic response of high-speed railway bridges. For this, based on measured responses from actual railway bridges, a numerical model was constructed using a numerical model updating technique. The dynamic response of the updated numerical model exhibited a strong agreement with the measured response from the actual railway bridges. Subsequently, this updated model was utilized to analyze the dynamic response characteristics of the bridges when KTX and EMU-320 trains operate at increased speeds. The maximum vertical displacement and acceleration at the mid-span of the bridges were also compared to those specified in the railway design standard with the increasing speed of KTX and EMU-320.

• Journal of The Korean Association For Science Education
• /
• v.24 no.3
• /
• pp.442-458
• /
• 2004

This study was conducted to develop a plan for a large-scale implementation of the Girl Friendly Science Program based on the results of analysis and investigation of its current pilot implementation, Girl Friendly Science Program materials, which was first developed in 1999 with the support from Ministry of Gender Equality, consist of 1) five theme-based units that are specifically targeted individual students' unique ability, aptitude, and career choice, and 2) differentiated learning materials for 7th through 10th grade female students. All the materials are available at the homepage (http://tes.or.kr/gfsp.cgi) of 'Teachers for Exciting Science(the organization of science teachers in Seoul area)'. Since the materials are well organized by topic and grade level and presented in both Korean word process document and html format, anyone can easily access to the materials for their own instructional use. Ever since its launch the number of visitors to the homepage has been constantly increasing. The evaluation results of the current pilot implementation of the materials that targeted individual students' ability and aptitude showed that it scored high in terms of its alignment to the original purpose, content, level, and effectiveness to implement in classrooms. However, its evaluation scores were low in terms of the convenience for teachers to guide the materials, and its organization and operation. The results also showed a significant change in students' perception of science, and students' positive experiences of science through various interdisciplinary activities. On the other hand, the evaluation of students' experiences with the materials showed that students' assessment about an activity was largely depending on a success or failure of their experiences. Overall, students' evaluation of activities scores were low for simple activities such as cutting off or pasting papers. According to students' achievement test results, differences between pre and post test scores in the Affective Domain was statistically significant (p<0.05), but not in Inquiry Domain. Based on teachers observations, numerous schools where have run this program reported that students' abilities to cooperate, discuss, observe and reason with evidences were improved. In order to implement this program in a larger scale, it is critical to have a strong support of teachers and induce them to change their teaching strategy through building a community of teachers and developing ongoing teacher professional development programs. Finally, there still remain strong needs to develop more programs, and actively discover and train more domestic woman scientists and engineers and collaborate with them to develop more educational materials for girls in all ages.