• Proceedings of the Korean Society of Disaster Information Conference
• /
• 2022.10a
• /
• pp.347-348
• /
• 2022

본 연구에서는 무인이동체를 활용한 철도교량의 외관조사 점검을 보다 효율적이고 객관성 있게 수행하기 위하여 무인이동체를 통해 촬영된 이미지를 딥러닝 기반 분석기술을 활용하여 손상 자동으로 분석 하기위한 기술을 연구하였다. 철도교량의 외관 손상 중 균열, 콘크리트 박리·박락, 누수, 철근노출에 대한 손상 이미지를 추출하여 딥러닝 분석 모델을 생성하고 학습한 분석 모델을 적용한 시스템을 실제 현장에 적용 테스트를 수행하였으며 학습 구현된 분석모델의 검측 재현율을 검토한 결과 평균 95%이상의 감지성능을 검토할 수 있었다. 개발 제안된 자동손상분석 기술은 기존 육안점검 결과 대비 보다 객관적이고 정밀한 손상 검측이 가능하며 철도 유지관리 분야에서 무인이동체를 활용한 외관조사 업무를 수행함에 있어 기존 대비 객관적인 결과도출과 소요시간, 비용저감이 가능할 것으로 기대된다.

• Proceedings of the Korean Society of Disaster Information Conference
• /
• 2017.11a
• /
• pp.197-198
• /
• 2017

인공재해와 자연재해로 인해 발생하는 비정상 하중에 의해 국부손상이 발생된 철골 교량 구조물은 추가적인 2차 붕괴의 위험요소들을 내재하고 있어 신속한 전면 해체가 요구된다. 본 시공 사례는 건설실패와 태풍 및 지진으로 국부손상이 발생된 철골 트러스 구조의 교량의 긴급해체를 위해 발파해체 공법을 적용한 사례이다. 철골 부재의 절단을 위해 성형폭약이 필요하지만 현지에서 수급이 불가능한 상태이기 때문에 장약용기를 직접 제작하고 에멀젼 폭약을 충전하여 만든 성형폭약을 이용하여 발파해체에 적용하였다. 직접 제작한 성형폭약을 이용하여 발파해체한 결과 철골 부재가 정확히 절단되면서 교량의 중앙부가 수직자유낙하하고, 교량의 양 끝단은 지지부를 중심으로 회전낙하 하였다. 또한 존치 구조물 및 주변에 피해가 발생하지 않았으며, 발파 후 파쇄 상태는 매우 양호하였다. 이로 인해 직접 제작한 성형폭약의 절단 성능을 확인할 수 있었으며, 신속하고 안전하게 국부손상이 발생된 구조물을 해체하였다.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.8 no.3
• /
• pp.77-86
• /
• 2004

The retrofit priority of existing and retrofitted bridges is examined and compared to determine effectively the seismic retrofit method of bridges. For the retrofit prioritization of bridges a quantitative procedure is proposed firstly based on seismic damage probabilities and total failure cost due to the damage of seismic vulnerable components. Using the proposed procedure, the retrofit priority of four typical girder-type bridges is determined. In addition, the ranking indices of bridges retrofitted by steel jackets and cable restrainers are revaluated for comparing with the results of existing bridges. Application of retrofitting method can considerably decreases damage possibilities of retrofitted components but may increases those of adjacent vulnerable components. Therefore, the seismic retrofitting effects based on the global motions of existing and retrofitted bridges should be examined to determine efficiently the retrofitting method. For evaluating the retrofitting effects the ranking indices obtained from the proposed procedure is found to be utilized effectively.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.43 no.2
• /
• pp.165-173
• /
• 2023

The data management and the evaluation of defects in the bridge are generally conducted based on inspection and diagnosis data, including the exterior damage map and defect quantity table prepared by periodic inspection. Since most of these data are written in 2D-based documents and are difficult to digitize in a standardized manner, it is challenging to utilize them beyond the defined functionality. This study proposed methods to efficiently build a BIM (Building Information Modeling)-based bridge damage model from raw data of inspection report and to manage and utilize the damage information linking to bridge model through the spread sheet data generated by COBie (Construction Operations Building Information Exchange). In addition, a method to conduct the condition assessment of defects in bridge was proposed based on an automatic evaluation process using digitized bridge member and damage information. The proposed methods were tested using superstructure of PSC-I girder concrete bridge, and the efficiency and effectiveness of the methods were verified.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.40 no.2
• /
• pp.233-238
• /
• 2020

Most deep learning (DL) approaches for bridge damage localization based on a structural health monitoring system commonly use supervised learning-based DL models. The supervised learning-based DL model requires the response data obtained from sensors on the bridge and also the label which indicates the damaged state of the bridge. However, it is impractical to accurately obtain the label data in fields, thus, the supervised learning-based DL model has a limitation in that it is not easily applicable in practice. On the other hand, an unsupervised learning-based DL model has the merit of being able to train without label data. Considering this advantage, this study aims to propose and theoretically validate a damage localization approach for bridges using a variational autoencoder, a representative unsupervised learning-based DL network: as a result, this study indicated the feasibility of VAE for damage localization.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.26 no.6
• /
• pp.93-101
• /
• 2022

RC Slab bridges in Korea account for more than 70% of the total bridges for more than 20 years of service. As the number of aging structures increases, the importance of safety diagnosis and maintenance of structures increases. For highway bridges, cracks are a main cause of deck deterioration, which is very closely related to the decrease in bridge durability and service life. In addition, the damage rate of expansion joints and bearings accounts for approximately 73% higher than that of major members. Therefore, this study defined damage scenarios combined with devices damages and deck deterioration. The stress distribution and maximum stress on the deck were then evaluated using design vehicle load and daily temperature gradient for single and combined damage scenarios. Furthermore, this study performed damage-spread analysis and predicted condition ratings according to a deck deterioration model generated from the inspection and diagnosis history data of cracks. The heterogeneous damages combined with the member damages of expansion joints and bearings increased the rate of crack area and damage spread, which accelerated the time to reach the condition rating of C. Therefore, damage to bridge members requires proper and prompt repair and replacement, and otherwise it can cause the damage to bridge deck and the spread of the damage.

• Journal of Korean Society of Steel Construction
• /
• v.22 no.1
• /
• pp.13-20
• /
• 2010

Seismic risk assessment of railway bridges is an important issue for a transportation network, because loss of functionality of railway bridges could result in severe disruption of the railway line, as no redundant routing systems generally exist. Although many studies have been conducted by numerous researchers regarding fragility analyses of bridge structure, little or no studies have been done for fragility analyses of a class of bridge structures considering their geometric variability. This study performs a fragility analysis for Track-on Steel-Plate-Girder (TOSPG) railway bridges in Korea considering their span variability. Seismic fragility curves are developed for a series of bridges with different spans varying from 2 to 15. At last, the fragility curves for the whole TOSPG bridges in Korea are also developed using the total probability theorem. This study is expected to effectively contribute to the seismic risk assessment of railway lines, where a number of bridges are present.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.42 no.1
• /
• pp.117-126
• /
• 2022

Currently, the number of bridge structures in Korea is continuously increasing and enlarged, and the number of old bridges that have been in service for more than 30 years is also steadily increasing. Bridge aging is being treated as a serious social problem not only in Korea but also around the world, and the existing manpower-centered inspection method is revealing its limitations. Recently, various bridge damage detection studies using deep learning-based image processing algorithms have been conducted, but due to the limitations of the bridge damage data set, most of the bridge damage detection studies are mainly limited to one type of crack, which is also based on a close set classification model. As a detection method, when applied to an actual bridge image, a serious misrecognition problem may occur due to input images of an unknown class such as a background or other objects. In this study, five types of bridge damage including crack were defined and a data set was built, trained as a deep learning model, and an open set recognition-based bridge multiple damage recognition model applied with OpenMax algorithm was constructed. And after performing classification and recognition performance evaluation on the open set including untrained images, the results were analyzed.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.21 no.5
• /
• pp.67-73
• /
• 2017

The concrete pavement slabs that suffer expansion due to the Alkali-Aggregate Reaction(AAR) increase and the increase consequently causes unexpected displacement of bridge abutment. As the expansion due to the AAR is greater than that due to the temperature change, lethal load can act on bridge abutment. Therefore appropriate preventive measures may be necessary. The degree of expansion by AAR depends on the severity of AAR and geometry condition of concrete pavement and road structure. In order to prevent damage to bridge, it is effective to release the expansion force of the concrete. It would be advantageous to replace the concrete pavement with asphalt for a long section of concrete pavement.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.38 no.4
• /
• pp.621-625
• /
• 2018

This paper proposes a method for bridge damage factor recognition from inspection reports using deep learning. Bridge inspection reports contains inspection results including identified damages and causal analysis results. However, collecting such information from inspection reports manually is limited due to their considerable amount. Therefore, this paper proposes a model for recognizing bridge damage factor from inspection reports applying Named Entity Recognition (NER) using deep learning. Named Entity Recognition, Word Embedding, Recurrent Neural Network, one of deep learning methods, were applied to construct the proposed model. Experimental results showed that the proposed model has abilities to 1) recognize damage and damage factor included in a training data, 2) distinguish a specific word as a damage or a damage factor, depending on its context, and 3) recognize new damage words not included in a training data.