• 한국의학물리학회지:의학물리
• /
• 제31권3호
• /
• pp.111-123
• /
• 2020

Deep learning (DL) is a subset of machine learning and artificial intelligence that has a deep neural network with a structure similar to the human neural system and has been trained using big data. DL narrows the gap between data acquisition and meaningful interpretation without explicit programming. It has so far outperformed most classification and regression methods and can automatically learn data representations for specific tasks. The application areas of DL in radiation oncology include classification, semantic segmentation, object detection, image translation and generation, and image captioning. This article tries to understand what is the potential role of DL and what can be more achieved by utilizing it in radiation oncology. With the advances in DL, various studies contributing to the development of radiation oncology were investigated comprehensively. In this article, the radiation treatment process was divided into six consecutive stages as follows: patient assessment, simulation, target and organs-at-risk segmentation, treatment planning, quality assurance, and beam delivery in terms of workflow. Studies using DL were classified and organized according to each radiation treatment process. State-of-the-art studies were identified, and the clinical utilities of those researches were examined. The DL model could provide faster and more accurate solutions to problems faced by oncologists. While the effect of a data-driven approach on improving the quality of care for cancer patients is evidently clear, implementing these methods will require cultural changes at both the professional and institutional levels. We believe this paper will serve as a guide for both clinicians and medical physicists on issues that need to be addressed in time.

• 한국정보통신학회논문지
• /
• 제24권4호
• /
• pp.467-473
• /
• 2020

최근 우리나라의 자동차 보급률이 증가함에 따라 교통사고 발생 건수 또한 증가하고 있다. 특히, 차량간 사고뿐만 아니라 횡단보도 근처에서의 인명 사고 또한 증가하고 있어 횡단보도 교통안전에 대한 주의가 더욱 요구되고 있다. 본 논문에서는 NVIDIA Jetson Nano급의 엣지 디바이스를 이용하여 횡단보도에 접근하는 차량을 인식하고 속도를 추정함으로써 횡단보도 주위 안전 상태를 예측하는 시스템을 제안한다. 딥러닝 기반 차량 위치 인식을 통하여 얻은 정보들을 바탕으로 다양한 기계 학습 기법을 학습시켜 차량 속도에 따른 위험 정도를 예측한다. 마지막으로, 실제 주행 영상을 이용한 실험 및 웹 시뮬레이션을 통해 제안하는 시스템의 성능과 활용 가능성을 검증하였다.

• 융합보안논문지
• /
• 제20권5호
• /
• pp.65-74
• /
• 2020

최근 확장현실(XR) 기술의 비약적인 발전으로 확장현실(XR) 기술을 활용한 교육훈련체계 개발 및 활용이 크게 늘고 있다. 특히 군사훈련 및 대테러 훈련과 같이 큰 위험성을 동반하고 많은 비용이 발생하는 분야에서는 확장현실(XR) 기술을 활용해 실제 훈련과 유사한 훈련환경을 조성함으로써 훈련성과는 높이고, 훈련에 소요되는 비용은 줄이며, 실제훈련에서 발생할 수 있는 안전사고의 위험을 최소화할 수 있는 확장현실(XR) 기반 모의훈련체계가 대안으로 떠오르고 있다. 본 논문에서는 육군사관학교 주관의 콘소시움에서 기(旣)개발한 5G 기반 증강현실(AR) 통합 지휘통제플랫폼을 기반으로 대한민국 경찰의 테러상황 대응 능력을 향상시키기 위한 확장현실(XR) 기반 초실감 대테러 교육훈련체계 구축방안 및 효과 검증방안을 제안한다. 제안하는 체계는 초실감 콘텐츠 및 훈련 시나리오를 바탕으로 대테러 능력향상을 위한 개인 및 팀 단위 가상훈련이 가능하고, 나아가 모의훈련장 및 실제 작전현장과 연계하여 현장 지휘통제소로도 즉시 활용이 가능하다.

• Geomechanics and Engineering
• /
• 제24권5호
• /
• pp.479-490
• /
• 2021

Pre-drainage of groundwater in the roof aquifer by boreholes is the main method for prevention of roof water disaster, and the drop in the water level during the drainage leads to the variation of the local stress in the overlying strata. Based on a multitude of boreholes for groundwater drainage from aquifer above the 1303 mining face of Longyun Coal Mine, theoretical analysis and numerical simulation are used to investigate the local stress variation in the process of borehole drainage. The results show that due to the drop in the water level of the roof aquifer during the drainage, the stress around the borehole gradually evolved. From the center of the borehole to the outside, a stress-relaxed zone, a stress-elevated zone, and a stress-recovered zone are sequentially formed. Along with the expansion of drainage influence, the stress peak in the stress-elevated zone also moves to the outside. When the radius of influence develops to the maximum, the stress peak position no longer moves outward. When the coal mining face advances to the drainage influence range, the abutment pressure in front of the mining face is superimposed with the high local stress around the borehole, which increases the risk of stress concentration. The present study provides a reference for the stress concentration caused by borehole drainage, which can be potentially utilized in the optimal arrangement of drainage boreholes in underground mining.

• 스마트미디어저널
• /
• 제11권5호
• /
• pp.56-62
• /
• 2022

배수지는 정수처리 된 물을 급수하기 위해 정수물을 모아두는 저장소로서, 물의 수요량에 따라 급수량을 조절하여 안정적으로 물을 공급하기 위해 배수지의 수위 관리는 매우 중요하다. 현재 배수지 내에 수위 계측 센서를 설치하여, 가압장의 펌프운영을 통해 배수지의 최적 수위를 관리하고 있으나, 센서의 오작동 및 통신두절 등 사고대응을 관리자 감시에 의존하고 있어, 사고의 위험을 안고 있다. 본 연구에서는 배수시설의 안정적 운영을 위하여, 배수지의 수위 변화 예측 인공지능 모델을 제안하였으며, 배수지 수위 변화 예측모델의 현장적용에 대한 안정성을 확인하기 위하여 수위 데이터의 결측 상황에 대한 시뮬레이션을 통하여, 실제 수위 변화값과 예측된 수위 변화값의 비교를 통하여 모델의 유용성을 확인하였다.

• 한국항공우주학회지
• /
• 제50권2호
• /
• pp.103-110
• /
• 2022

악천후나 기계적 결함에 의한 비행성(Handling Qualities) 저하는 그 상황에 익숙하지 못한 조종사에게 치명적인 위험을 줄 수 있다. 특히 결빙의 경우에는 실제로 사고가 빈번하게 일어나는 원인인 만큼 중요하게 고려해야 할 사안이다. 대다수의 기존 연구들은 결빙에 따른 공기역학적 성능 변화와 그에 따른 모델링 방법을 제시하거나 결빙을 방지하는 방법을 주로 다루었으나, 본 연구에서는 전기체(full-scale) 헬리콥터에 발생하는 결빙으로 인한 비행 성능 저하를 능동적으로 보상하는 비행제어기를 설계하고자 한다. 본 연구에서는 먼저 UH-60 헬리콥터에 발생하는 결빙으로 인한 비행성 저하를 CONDUIT이라는 프로그램을 통해 보이고, 이러한 비행성 저하를 보상하기 위해 RS-LQR(Robust Servomechanism Linear Quadratic Regulation) 기법을 사용한 강인 제어기를 설계하였다. 시뮬레이션을 통해 제안한 강인 제어기가 헬리콥터가 결빙된 상황에서도 Level 1 비행성을 유지하는 것을 보였다.

• 한국지열·수열에너지학회논문집
• /
• 제18권3호
• /
• pp.7-18
• /
• 2022

Secondary batteries used in electric vehicles have a potential risk of ignition and explosion. Various safety measures are being taken to prevent these risks. A numerical study was performed using a computational fluid dynamics code on the cases where pressure relief vents that can reduce the blast overpressures of batteries were installed in the through-compression test room, short-circuit drop test room, combustion test room, and immersion test room in facilities rleated to battery used in electric vehicles. This study was conducted using the weight of TNT equivalent to the energy release from the battery, where the the thermal runaway energy was set to 324,000 kJ for the capacity of the lithium-ion battery was 90 kWh and the state of charge (SOC) of the battery of 100%. The explosion energy of TNT (△H_TNT) generally has a range of 4,437 to 4,765 kJ/kg, and a value of 4,500 kJ/kg was thus used in this study. The dimensionless explosion efficiency coefficient was defined as 15% assuming the most unfavorable condition, and the TNT equivalent mass was calculated to be 11 kg. The internal explosion generated in a test room shows the very complex propagation behavior of blast waves. The shock wave generated after the explosion creates reflected shock waves on all inner surfaces. If the internally reflected shock waves are not effectively released to the outside, the overpressures inside are increased or maintained due to the continuous reflection and superposition from the inside for a long time. Blast simulations for internal explosion targeting four test rooms with pressure relief vents installed were herein conducted. It was found that that the maximum blast overpressure of 34.69 bar occurred on the rear wall of the immersion test room, and the smallest blast overpressure was calculated to be 3.58 bar on the side wall of the short-circuit drop test room.

• Geomechanics and Engineering
• /
• 제28권3호
• /
• pp.283-298
• /
• 2022

Landslides triggered by the combination of heavy precipitation and anthropological disturbance in hilly areas cause severe damage to human lives, properties, and infrastructure constructions. A comprehensive investigation of the influencing factors and failure mechanisms of landslides are significant for disaster mitigation and prevention. This paper utilized the combination of detailed geological investigation, physical experimental testing as well as numerical modelling to determine the failure mechanism, and proposed a countermeasures of the Lantian landslide occurred on 2, July 2017. The results reveal that the Lantian landslide is a catastrophic reactivated slide which occurred in an active tectonic region in Southwest China. Because of the unique geological settings, the fully to highly weathered basalts in the study area with well-developed fractures favored the rainwater infiltration, which is the beneficial to slide reactivation. Engineering excavation and heavy precipitation are the main triggering factors to activate the slide motion. Two failure stages have been identified in the landslide. The first phase involves a shallow mass collapse originated at the upper slopes, which extends from the road to platform at rear part, which is triggered by excavation in the landslide region. Subjected to the following prolonged rainfall from 19 June to 2 July, 2017, the pore water pressure of the slope continually increased, and the groundwater table successively rise, resulting in a significant decrease of soil strength which leads to successive large-scale deep slide. Thereinto, the shallow collapse played a significant role in the formation of the deep slide. Based on the formation mechanisms of the landslide, detailed engineering mitigation measures, involving slope cutting, anchor cable frame, shotcrete and anchorage, retaining wall and intercepting ditch were suggested to reduce the future failure risk of the landslide.

• 방송공학회논문지
• /
• 제26권4호
• /
• pp.356-367
• /
• 2021

최근 사물인터넷 및 인공지능 기술의 발전에 따라 제조, 스마트시티 등 다양한 분야에서 실시간으로 데이터를 수집하고 분석하여 현실세계 문제에 대한 최적화를 수행하는 연구 및 적용사례가 증가하고 있다. 대표적으로 현실세계를 디지털화한 가상세계와 양방향으로 실시간 동기화를 지원하는 디지털 트윈 기술이 주목받고 있다. 본 논문에서는 디지털 트윈을 정의하고 사물인터넷 국제표준인 oneM2M 기반의 IoT 플랫폼을 활용하여 현실사물과 가상세계의 예측결과를 실시간으로 연결하는 디지털 트윈 플랫폼의 프로토타입을 제안한다. 또한, 제안된 프로토타입을 적용하여 물체의 충돌을 사전에 예측하여 사고를 예방할 수 있는 응용서비스를 구현한다. 응용서비스에서는 사전 정의한 테스트 케이스 수행을 통해 제안한 디지털 트윈 프로토타입이 크레인의 동작을 사전 예측하여 충돌 위험을 감지하고 이를 기반으로 최적 제어를 수행할 수 있으며 실제 환경에 응용 가능함을 보였다.

• Journal of Radiation Protection and Research
• /
• 제46권2호
• /
• pp.39-47
• /
• 2021

Background: The National Institutes for Quantum and Radiological Science and Technology-National Institute of Radiological Sciences (QST-NIRS) has continuously investigated the undesired radiation exposure in ion beam radiotherapy mainly in carbon-ion radiotherapy (CIRT). This review introduces our investigations on the secondary neutron dose in CIRT with the broad and scanning beam methods. Materials and Methods: The neutron ambient dose equivalents in CIRT are evaluated based on rem meter (WENDI-II) measurements. The out-of-field organ doses assuming prostate cancer and pediatric brain tumor treatments are also evaluated through the Monte Carlo simulation. This evaluation of the out-of-field dose includes contributions from secondary neutrons and secondary charged particles. Results and Discussion: The measurements of the neutron ambient dose equivalents at a 90#x00B0; angle to the beam axis in CIRT with the broad beam method show that the neutron dose per treatment dose in CIRT is lower than that in proton radiotherapy (PRT). For the scanning beam with the energy scanning technique, the neutron dose per treatment dose in CIRT is lower than that in PRT. Moreover, the out-of-field organ doses in CIRT decreased with distance to the target and are less than the lower bound in intensity-modulated radiotherapy (IMRT) shown in AAPM TG-158 (American Association of Physicists in Medicine Task Group). Conclusion: The evaluation of the out-of-field doses is important from the viewpoint of secondary cancer risk after radiotherapy. Secondary neutrons are the major source in CIRT, especially in the distant area from the target volume. However, the dose level in CIRT is similar or lower than that in PRT and IMRT, even if the contributions from all radiation species are included in the evaluation.