• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.104-106
• /
• 2021

도시의 인구 밀집도가 증가함에 따라 도시의 단위 면적당 건물 밀집도 역시 증가하고 있으며, 이에 도시 화재는 대규모 화재로 발전할 가능성이 높다. 도시 내 대규모 화재로 인한 인명 및 경제적인 피해를 최소화하기 위해 시뮬레이션 기반의 화재 대응 방안들이 널리 연구되고 있으며, 최근에는 시뮬레이션에서 효과적인 화재 대응 방안을 탐색하기 위해 강화학습 기술을 활용하는 연구들이 소개되고 있다. 그러나, 시뮬레이션의 규모가 커지는 경우, 상태 정보 및 화재 대응을 위한 행위 공간의 크기가 증가함으로 인해 강화학습의 복잡도가 증가하며, 이에 따라 학습 확장성이 저하되는 문제가 발생한다. 본 논문에서는 시뮬레이션 규모 증가 시 강화학습의 학습 확장성을 유지하기 위해, 화재 상황 정보와 재난 대응을 위한 행위 공간을 변환하는 기법을 제안한다. 실험 결과를 통해 기존에 강화학습 모델의 학습이 어려웠던 대규모 도시 재난시뮬레이션에서 본 기법을 적용한 강화학습 모델은 학습 수행이 가능하였으며, 화재 피해가 없는 상황의 적합도를 100%로 하고, 이것 대비 99.2%의 화재 대응 적합도를 달성했다.

• Proceedings of the KIEE Conference
• /
• summer
• /
• pp.521-523
• /
• 2004

산불과 같은 대형화재는 철탑의 전선이나 애자와 같은 부착물의 기능에 많은 장애를 유발할 수 있어 송전선로 운영에 큰 영향을 미치게 될 수 있다. 만약 이와 같은 재해로 인해 송전선로의 운영에 차질이 생긴다면 산업전반에 걸쳐 엄청난 파급효과를 초래시킬 것이 다. 본 논문에서는 산불과 같은 화재가 송전용 폴리머애자에 미치는 영향을 파악하기 위하여 산불의 불꽃을 모의한 실험장치를 제작하였으며, 시료로는 현재 송전선로에 사용되고 있는 송전용 폴리머애자를 축소 제작한 폴리머애자와 자기애자를 사용하였다. 산불모의 인공화염 실험은 가열시간의 경과에 따라 폴리머애자의 하우징과 자기애자의 디스크 변화를 관찰하였고, 가열실험이 끝난 애자들을 이용하여 전기적인 시험과 기계적인 시험을 행하였다. 이 실험 데이터를 바탕으로 폴리머애자와 자기애자의 특성변화를 비교 분석하여 산불영향이 송전용 폴리머애자에 미치는 영향을 연구하였다.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1998.10a
• /
• pp.522-526
• /
• 1998

본 논문은 가정용 배전반 시스템을 디지털식으로 구현하고, 기존의 디지털식 배전반 시스템에 없는 월 수요전력량 예측과 화재발생의 원인 중에 하나인 옥내 전선선로의 결함을 신경회로망으로 검출하여 차단하는 인공지능형 가정용 배전반 시스템을 하드웨어로 구현하고 실험하였으며, 그 결과를 제시하였다.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.12 no.2
• /
• pp.331-336
• /
• 2017

The need for early fire detection technology is increasing in order to prevent fire disasters. Sensor device detection for heat, smoke and fire is widely used to detect flame and smoke, but this system is limited by the factors of the sensor environment. To solve these problems, many image-based fire detection systems are being developed. In this paper, we implemented a system to detect fire and smoke from camera input images using a convolution neural network. Through the implemented system using the convolution neural network, a feature map is generated for the smoke image and the fire image, and learning for classifying the smoke and fire is performed on the generated feature map. Experimental results on various images show excellent effects for classifying smoke and fire.

• Economic and Environmental Geology
• /
• v.53 no.3
• /
• pp.287-296
• /
• 2020

Mine Reclamation Project is being carried out with the aim of ensuring a sustainable green living and helping to develop eco-friendly mines by analyzing, removing and preventing the harmful factors. Mines developed during the japanese colonial period and mining boom period are still not repaired throughout the country, and from these scattered risks, public safety is worth pursuing as a top priority. The project that is close to public safety in the mine recalmation project is an emergency treatment, and the most widely used method is a filling method similar to the ground subsidence prevention. If dangerous mine cavity or tunnels are located in the mountains, charging with existing materials may not be possible, or unreasonable cases may occur, and new methods of technological development are required. Emergency actions should be carried out safely and efficiently to prevent the loss of precious people's lives on the hiking paths adjacent to dangerous mining sites. In these field conditions, urethane foam materials may be an alternative. In this study, the applicability of urethane foam materials in mining was reviewed through overseas cases. It was also tested on the appropriate depth of top soil for the protection of urethane foam materials through forest fire simulation test. The test result show that approximately 15cm of soil covering (recommended 20cm over) was suitable for maintaining the function of foam materials from forest fires.

• Fire Science and Engineering
• /
• v.25 no.2
• /
• pp.107-113
• /
• 2011

This paper presents the results of artificially accelerated aging characteristics of the rate of rise spot type heat detectors. This experiment carried out to investigate the detector's operating characteristics when a fire was broken out or the regular tests were performed. The result showed that the delay of operating time or no-operation of heat detector which was made by company B and used in the field for 5 years may be occurred even at the $100^{\circ}C$. This result is due to the leakage of inflated air from heat chamber. However the operating display LED was not out of order, even if the temperature was increased up to $160^{\circ}C$.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.27 no.5
• /
• pp.675-682
• /
• 2021

Fire accidents on land and at sea can cause serious casualties; specifically, owing to the nature of marine plants and ships, the mortality rate at sea from suffocation in confined spaces is significantly higher than that on land. To prevent such cases of asphyxiation, it is essential to install ventilation fans that can outwardly direct these toxic gases from fires; however, considering the scale of marine fires, the installation of large ventilation fans is not easy owing to the nature of marine structures. Therefore, in this study, we developed a new concept for fire safety technology to control toxic gases generated by fires from applied direct current (DC) electric fields. In the event of a fire, most flames contain large numbers of positive and negative charges from chemi-ionization, which generates an "ionic wind" by Lorentz forces through the applied electric fields. Using these ionic winds, an experimental study was performed to artificially control the fire smoke caused by burning paper and styrofoam, which are commonly used as insulation materials in general buildings and ships. The experiments showed that a fire smoke could be artificially controlled by applying a DC voltage in excess of ±5 kV and that relatively effective control was possible by applying a negative voltage rather than a positive voltage.

• 재활복지
• /
• v.21 no.3
• /
• pp.129-146
• /
• 2017

In an event of an emergency, such as fire in a building, visually impaired and blind people are prone to exposed to a level of danger that is greater than that of normal people, for they cannot be aware of it quickly. Current fire detection methods such as smoke detector is very slow and unreliable because it usually uses chemical sensor based technology to detect fire particles. But by using vision sensor instead, fire can be proven to be detected much faster as we show in our experiments. Previous studies have applied various image processing and machine learning techniques to detect fire, but they usually don't work very well because these techniques require hand-crafted features that do not generalize well to various scenarios. But with the help of recent advancement in the field of deep learning, this research can be conducted to help solve this problem by using deep learning-based object detector that can detect fire using images from security camera. Deep learning based approach can learn features automatically so they can usually generalize well to various scenes. In order to ensure maximum capacity, we applied the latest technologies in the field of computer vision such as YOLO detector in order to solve this task. Considering the trade-off between recall vs. complexity, we introduced two convolutional neural networks with slightly different model's complexity to detect fire at different recall rate. Both models can detect fire at 99% average precision, but one model has 76% recall at 30 FPS while another has 61% recall at 50 FPS. We also compare our model memory consumption with each other and show our models robustness by testing on various real-world scenarios.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.12
• /
• pp.63-71
• /
• 2020

In this paper, we propose a fire detection technology using YOLOv3 and EfficientDet, the most reliable artificial intelligence detection algorithm recently, an alert service that simultaneously transmits four kinds of notifications: text, web, app and e-mail, and an AWS system that links fire detection and notification service. There are two types of our highly accurate fire detection algorithms; the fire detection model based on YOLOv3, which operates locally, used more than 2000 fire data and learned through data augmentation, and the EfficientDet, which operates in the cloud, has conducted transfer learning on the pretrained model. Four types of notification services were established using AWS service and FCM service; in the case of the web, app, and mail, notifications were received immediately after notification transmission, and in the case of the text messaging system through the base station, the delay time was fast enough within one second. We proved the accuracy of our fire detection technology through fire detection experiments using the fire video, and we also measured the time of fire detection and notification service to check detecting time and notification time. Our AI fire detection and notification service system in this paper is expected to be more accurate and faster than past fire detection systems, which will greatly help secure golden time in the event of fire accidents.

• Journal of the Korea Society for Simulation
• /
• v.19 no.2
• /
• pp.1-8
• /
• 2010

With recent development of computer hardware and 3D graphic technique, a lot of people have concern for something to express as the 3D graphic that look the real environment. Because the request of users have increased, the 3D simulation is developed and popularized in the many field. In this paper, we design and implement the simulation system that humans evacuate a building fires using the 3D graphic techniques. In this paper, we use the A* algorithm to humans have the artificial intelligence at evacuating a building fires, calculate the evacuation speed of each human considering temperature damage and smoke damage. In this paper, we applied the real building to demonstrate the effect of proposed evacuation simulation. Experimental results showed that the evacuation speed is affected by the temperature condition and the smoke density.