• Journal of the Korean Society of Marine Environment & Safety
• /
• v.28 no.7
• /
• pp.1120-1128
• /
• 2022

In the event of a marine accident, the longer the exposure time to the sea increases, the faster the chance of survival decreases. However, because the search area of the sea is extremely wide compared to that of land, marine object detection technology based on the sensor mounted on a satellite or an aircraft must be applied rather than ship for an efficient search. The purpose of this study was to rapidly detect an object in the ocean using a hyperspectral image sensor mounted on an aircraft. The image captured by this sensor has a spatial resolution of 8,241 × 1,024, and is a large-capacity data comprising 127 spectra and a resolution of 0.7 m per pixel. In this study, a marine object detection model was developed that combines a seawater identification algorithm using DBSCAN and a density-based land removal algorithm to rapidly analyze large data. When the developed detection model was applied to the hyperspectral image, the performance of analyzing a sea area of about 5 km² within 100 s was confirmed. In addition, to evaluate the detection accuracy of the developed model, hyperspectral images of the Mokpo, Gunsan, and Yeosu regions were taken using an aircraft. As a result, ships in the experimental image could be detected with an accuracy of 90 %. The technology developed in this study is expected to be utilized as important information to support the search and rescue activities of small ships and human life.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2015.07a
• /
• pp.239-241
• /
• 2015

인적오류는 해양사고의 주요한 원인을 차지하고 있으나 그동안 이에 대한 연구가 미흡한 실정이다. 본 연구에서는 해양사고의 인적요소 중 혼잡해역이나 협수로 등에서 타 선박에 대한 다양한 조우 상황별 항해사가 받는 스트레스 값을 실험하였다. 또한 GEV distribution을 적용하여 인적오류 평가모델을 개발하고자 하였고 선박 실험을 통하여 도출된 데이터를 통하여 분포함수와 스트레스 수치 사이의 오차를 계산하고 평가하였다. 향후 본 모델의 개발을 통하여 선박 운항자의 스트레스 정도를 예측 할 수 있고 이에 따른 사고 예방 조치 및 적절한 교육훈련 실시 등 다양한 인적 오류 저감을 위한 대책마련에 도움이 되고자 하는 목적이다.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.27 no.7
• /
• pp.1088-1097
• /
• 2021

Vibration data of mechanical equipment inevitably have noise. This noise adversely af ects the maintenance of mechanical equipment. Accordingly, the performance of a learning model depends on how effectively the noise of the data is removed. In this study, the noise of the data was removed using the Denoising Auto Encoder (DAE) technique which does not include the characteristic extraction process in preprocessing time series data. In addition, the performance was compared with that of the Wavelet Transform, which is widely used for machine signal processing. The performance comparison was conducted by calculating the failure detection rate. For a more accurate comparison, a classification performance evaluation criterion, the F-1 Score, was calculated. Failure data were detected using the One-Class SVM technique. The performance comparison, revealed that the DAE technique performed better than the Wavelet Transform technique in terms of failure diagnosis and error rate.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.2
• /
• pp.235-241
• /
• 2022

Several studies using Convolutional Neural Network (CNN) have been conducted to detect and classify the sounds of marine mammals in underwater acoustic data collected through passive acoustic monitoring. In this study, the possibility of automatic classification of bearded seal sounds was confirmed using a CNN model based on the underwater acoustic spectrogram images collected from August 2017 to August 2018 in East Siberian Sea. When only the clear seal sound was used as training dataset, overfitting due to memorization was occurred. By evaluating the entire training data by replacing some training data with data containing noise, it was confirmed that overfitting was prevented as the model was generalized more than before with accuracy (0.9743), precision (0.9783), recall (0.9520). As a result, the performance of the classification model for bearded seals signal has improved when the noise was included in the training data.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.27 no.1
• /
• pp.60-66
• /
• 2021

The importance of maintenance is leading the application of the maintenance strategy in various industries. The maritime industry is also a part of them, with changes in selecting and applying the maintenance strategy, but rather slowly, by retaining the old strategy. In particular, the ship is maintaining a previously used strategy. In the circumstance of the sea, ship requires a new suggestion for maintenance strategy. A ship predictive maintenance model predicts the breakdown or malfunction of machineries to secure maintenance time with preventive actions and treatments, thereby avoiding maintenance-related dangerous factors. This study focused on applying text analysis to an Liquefied Natural Gas Carriers dock indent document, and the analysis results were interpreted from the original document. The inter-relational patterns observed from the frequency of common maintenance combinations among different parts and equipment in ships will be applied to the development of ship predictive maintenance.

• Journal of Navigation and Port Research
• /
• v.43 no.3
• /
• pp.160-165
• /
• 2019

The conduct of seafarer is major cause of marine accidents. This study models the behavior of the seafarer based on the Hidden Markov Model (HMM). Additionally, through the path analysis of the behavior estimated by the model, the kind of situations, procedures and errors that may have caused the marine accidents were interpreted. To successfully implement the model, the seafarer behaviors were observed by means of the summarized verdict reports issued by the Korean Maritime Safety Tribunal, and the observed results converted into behavior data suitable for HMM learning through the behavior classification framework based on the SRKBB (Skill-, Rule-, and Knowledge-Based Behavior). As a result of modeling the seafarer behaviors by the type of vessels, it was established that there was a difference between the models, and the possibility of identifying the preferred path of the seafarer behaviors. Through these results, it is expected that the model implementation technique proposed in this study can be applied to the prediction of the behavior of the seafarer as well as contribute to the prioritization of the behavior correction among seafarers, which is necessary for the prevention of marine accidents.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2015.10a
• /
• pp.202-204
• /
• 2015

2010년부터 최근 5년간 해양안전심판원의 충돌사고 통계분석 결과에 따르면 98%가 인적과실이 원인이 될 정도로 경계소홀 등 해기사에 의한 운항과실이 해양사고의 대부분을 차지하고 있으나 그동안 외국뿐만 아니라 국내에서도 이에 대한 연구는 극히 미흡한 실정이다. 본 연구는 항해중인 선박이 타 선박과 6가지 유형(Head-on, $045^{\circ}$, $090^{\circ}$, $135^{\circ}$, Overtaking, Overtaken)의 다른 방위각으로 조우 상황시 양 선박의 거리가 가까워지면서 거리별 항해당직사관인 해기사가 느끼는 스트레스(주관적 위험도)를 실제 항해중인 선박에 승선하여 실험하였고 획득한 데이터를 통계적으로 분석 및 처리하였다. 데이터를 곡선적합(Curve fitting) 적용 결과 6가지 조우각도별 거리에 따른 연속 데이터 획득이 가능 하였고 인적과실의 주요 원인 중의 하나인 타 선박 조우 상황시 해기사가 느끼는 스트레스 일반화 모델을 위한 기초 식을 제안하였다. 이러한 결과는 향후 해기사의 인적과실을 줄여 해양사고 예방을 위한 기초 자료로 활용 할 수 있을 것이다.

• Proceedings of KOSOMES biannual meeting
• /
• 2017.11a
• /
• pp.162-163
• /
• 2017

해양 HNS(Hazardous and Noxious Substances)의 유출 사고 시, 막대한 인명 피해와 환경 훼손을 피하기 위해 유출 사고 조기 예측과 정확한 확산 경로를 예측하는 것이 필수적이다. 본 연구의 최종목적은 전산유체역학을 이용하여 HNS사고가 발생하였을 때 위험구역을 적절히 예측할 수 있는 수치해석기법을 개발하고, 다양한 해양사고조건과 환경영향을 고려하여 근접역에서의 2차원 확산 특성을 고찰하고 확산 현상을 예측하기 위한 모델을 개발하는 것이다. 본 연구에서는 상용코드인 ANSYS FLUENT(V. 17.2)을 사용하여 근접역에서의 2차원 확산특성을 모사하고 분석하였다. 특히, 누출된 HNS의 위치별 농도를 예측하기 위해 종수송방정식(Species Transport Equation)을 이용하였으며 RANS(Reynolds-Averaged Navier-Stokes) 방정식과 표준 $k-{\varepsilon}$ 모델을 이용하여 난류유동을 모사하였다. 해석된 결과는 문헌에서 얻어진 실험데이터와 상호비교하였으며 해수의 유속, HNS의 밀도에 따른 유층 두께, 해수면 HNS 평균 농도 그리고 HNS 전파 속도를 분석하였다. 유층 두께는 해류 유속에 따라 변화하며 변화 경향에 따라 두 구간으로 나눌 수 있다. 해류 전파 속도는 대체로 해류 유속과 선형적 비례관계를 갖는 것으로 나타났다. 해수면 평균 HNS 농도는 해류 유속에 선형적으로 비례하여 감소하며, HNS 밀도가 큰 경우 해수면 평균 HNS 체적 농도는 더 빠르게 감소하게 된다. 이러한 결과는 HNS 확산 특성을 분석하고 관련된 예측모델을 개발하는 데에 기여할 수 있다.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2023.05a
• /
• pp.210-211
• /
• 2023

To reduce maritime accidents and improve safety for domestic ships, the government has developed and is operating the Korean e-Navigation Service. The ship collision and grounding monitoring service and the route planning service were developed based on the experience of captains and navigators who navigated large ships. However, in Korea, small ships account for a large portion of the users of these services, so there are various difficulties and additional requirements for applying services based on large ships to small ships. To improve this, a data science-based algorithm model using maritime digital traffic information is being developed for small ships, and it is important to evaluate the performance of the developed model and compare its accuracy with existing services. In this study, we propose a performance evaluation method for the developed algorithm model and study how to compare and evaluate its accuracy with existing e-navigation services.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2021.11a
• /
• pp.5-7
• /
• 2021

최근 해양사고 발생 건수의 급격한 증가와 더불어 어선의 조업 중 안전사고로 인한 인명피해 또한 크게 증가하였다. 중앙해양안전심판원의 공식 통계에 따르면, 2017년 46명이었던 안전사고의 사망실종자는 2019년 38명으로 소폭 감소하였으나, 2020년 60명으로 크게 증가하였다. 그러나, 사망자가 감소하였던 2019년 안전사고로 인한 부상자는 164명으로 전년도 76명 대비 2배 이상 증가하며 어선원에 대한 안전재해 예방은 실효성을 갖지 못하는 실정이다. 국내 업종별 산업재해율을 비교해볼 때, 어업 재해율은 농업, 광업, 제조업, 건설업, 임업 등을 포괄한 전체 산업 평균 재해율의 약 10배에 이르며 어업인들의 안전이 큰 위협에 놓여있음을 시사한다. 본 연구에서는 2017년부터 2019년의 수협중앙회의 어선원 공제보험데이터를 활용하여 선박별, 재해자별 사고 현황과 발생 형태를 분석하였다. 특히, 교차분석과 연관규칙분석기법을 통해 승선 직책별 부상 부위와 사고발생 형태를 식별하였으며, 이에 따라 직책에 따른 부상 부위를 비교하여 맞춤형 예방대책 수립을 위한 지원과, 사고발생형태의 군집 분석을 통해 발생형태간의 연결고리를 도출하여, 스위스 치즈 모델에서 제안하는 취약점(Weakness)를 식별하고, 이러한 취약점을 보완하기 위한 방어 장벽(Protective barriers)을 제언한다.