• Proceedings of the KIPE Conference
• /
• 2019.07a
• /
• pp.537-539
• /
• 2019

Recently, large-scale fires have been generated as urban buildings have become more and more density. Especially, the expansion of smoke in buildings due to high-rise is an problem, and the smoke is the main cause of death in fires. Therefore, in this paper, the image-based smoke detection is proposed through deep learning-based artificial intelligence techniques to prevent possible damage if existing detectors are not detected. In addition, the detection model was not configured simply through only the smoke data set, but the data set in the haze form was additionally composed together to compensate for the accuracy.

• International Journal of Advanced Culture Technology
• /
• v.10 no.4
• /
• pp.536-546
• /
• 2022

A society will lose a lot of something in this field when the forest fire broke out. If a forest fire can be detected in advance, damage caused by the spread of forest fires can be prevented early. So, we studied how to detect forest fires using CCTV currently installed. In this paper, we present a deep learning-based model through efficient image data construction for monitoring forest fire smoke, which is unstructured data, based on the deep learning model YOLOv5. Through this study, we conducted a study to accurately detect forest fire smoke, one of the amorphous objects of various forms, in YOLOv5. In this paper, we introduce a method of self-learning by producing insufficient data on its own to increase accuracy for unstructured object recognition. The method presented in this paper constructs a dataset with a fixed labelling position for images containing objects that can be extracted from the original image, through the original image and a model that learned from it. In addition, by training the deep learning model, the performance(mAP) was improved, and the errors occurred by detecting objects other than the learning object were reduced, compared to the model in which only the original image was learned.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.13 no.6
• /
• pp.451-462
• /
• 2011

The smoke control system plays the most important role in securing evacuation environment when a fire occurs in road tunnels. Smoke control methods in road tunnels are classified into two categories which are longitudinal ventilation system and transverse ventilation system. In this study it is intended to review the characteristics of smoke behavior by performing numerical analysis for calculating the optimal smoke exhaust air volume with scaled-model and simulation when a fire occurs in tunnels in which transverse ventilation is applied, and for obtaining the basic data required for the design of smoke exhaust systems by deriving optimal smoke exhaust operational conditions for various conditions. As a result of this study, when the critical velocity in the tunnel is 1.75 m/s and 2.5 m/s, the optimal smoke exhaust air volume has to be more than $173m^3/s$, $236m^3/s$ for the distance of the smoke moving which can limit the distance to 250 m. In addition, in case of uniform exhaust the generated smoke is effectively taken away if the two exhaust holes near the fire region are opened at the same time.

• Proceedings of the KSR Conference
• /
• 2007.05a
• /
• pp.1319-1324
• /
• 2007

Many researches have been performed to analyze the smoke movement in tunnel fires by using field model. Recently, FDS(Fire Dynamics Simulator) v.4, which is one of the field model and developed from NIST(National Institute of Standards and Technology), is widely used. In tunnel fires, FDS can show detail results in local point, but it has difficulties in boundary condition and taking long computing time as the number of grid increases. So, there is a need to use alternative method for tunnel fire simulation. A zone model is different kind of CFD method and solves ordinary differential equation based on conservation and auxiliary equations. It shows good macroscopic view in less computing time compared to field model. In this study, therefore, to confirm the applicability of CFAST in tunnel fire analysis, numerical simulations using CFAST are conducted to analyze smoke movement in longitudinal ventilation reduced-scale tunnel fires. Then the results are compared with experimental results. The differences of temperature and critical velocity between numerical results and experimental data are over $30^{\circ}C$ and 0.9m/s, respectively. These values are out of error range. It shows that CFAST 6.0 is hard to be used for tunnel fire simulation.

• Tunnel and Underground Space
• /
• v.16 no.6 s.65
• /
• pp.526-535
• /
• 2006

An experimental study was carried out on a reduced scale tunnel model to grasp the behavioral feature of fire-induced smoke in the long tunnels. Based on Froude modeling, the 1/50 scaled tunnel model (20 m long) was constructed by acrylic tubes and paraffin gas was released inside the tunnel to simulate the 20 MW fire-induced smoke. me test results show, that after approximately 2 minutes of fire generation, was descended from the tunnel ceiling through the decrease of buoyancy, then it was symmetrically propagated about 90 meters for 4 minutes before jet fans were operated. The smoke was effectively controlled when the jet fans were operated and an air stream velocity was getting closed to reach a critical velocity (the minimum air velocity that requires to suppress the smoke spreading against the longitudinal ventilation flow during the tunnel fire situations). It was also found out that a range of smoke was spreaded about 3 meters from the origin of fire but the range was not propagated to the escape direction anymore. The early stage of the In operation, however, showed that the smoke was hardly controlled. It means that the operation of emergency ventilation system has many dangerous factors such as an intercepting breathing zone.

• Proceedings of the KSR Conference
• /
• 2008.11b
• /
• pp.66-72
• /
• 2008

In this study, transient 3D numerical simulations were performed to analyze the characteristics of fire driven flow for smoke ventilation system operating conditions in the deeply underground subway station. The smoke flow patterns were compared and discussed under smoke fan operating mode and off mode in the platform. Soongsil Univ. station(line number 7)was chosen for simulation which was the one of the deepest underground subway stations in the each lines of Seoul. The geometry for model is 365m in length include railway, 23.5m for width, 47m for depth. Therefore 10,000,000 structured grids were used for fire simulation. The parallel computational method for fast calculation was employed to compute the heat and mass transfer eqn's with 6 CPUs(Intel 3.0GHz Dual CPU, 12Cores) of the linux clustering machine. The fire driven flow was simulated with using FDS code in which LES method was applied. The Heat release rate was 10MW and The Ultrafast model was applied for the growing model of the fire source.

• 한국전산유체공학회:학술대회논문집
• /
• 2003.10a
• /
• pp.225-226
• /
• 2003

• Tunnel and Underground Space
• /
• v.12 no.1
• /
• pp.31-36
• /
• 2002

This paper concerns smoke propagation in tunnel fires with various size of fire source. Experiments carried out in model tunnel and those results were compared with numerical results. The Froude scaling law was used to scale model tests for comparison with larger scale tests. In order to validate for numerical analysis, temperature distribution of predicted data was compared with measured data. Examining the temperature distribution, we found that smoke layer does not come down under 50% of tunnel heights for a short tunnel heights for a short tunnel firs without ventilation. Front velocity of smoke layer is proportional to the cube root of heat release rate. And it is in good agreement with existing empirical expression and numerical prediction. In a short tunnel fire, horizontal propagation of smoke layer is more important than vertical smoke movement for evacuation plan.

• Fire Science and Engineering
• /
• v.30 no.1
• /
• pp.63-73
• /
• 2016

This study examined the effects of the wind conditions, such as wind velocity and wind directions, on the performance of the mechanical smoke exhaust systems for high-rise building fires. A scaled model design and CFD simulations were used to verify the effects both quantitatively and qualitatively. The results showed that the smoke exhaust velocity of the mechanical exhaust system can be reduced by up to 17% at a wind velocity of 5 m/s (equivalent to an outdoor wind speed of 16 m/s) and a wind direction of ${\theta}=5^{\circ}$. In addition, the angle of the outdoor wind direction below ${\theta}=25^{\circ}$ had a significantly influence on the smoke exhaust flow rate and reduced exhaust performance of the smoke exhaust system in a fire.

• Fire Science and Engineering
• /
• v.17 no.2
• /
• pp.56-61
• /
• 2003

The smoke removal rate from a room with an opening was investigated for different smoke control systems by using the large eddy simulation turbulence model of the Fire Dynamics Simulator. The decreasing rate of the particles randomly distributed in the 2m X 2m X 2.4m room was com-pared for the ventilation system, pressurization system and extraction system, and for the air flowrate of the ventilation system. Difference in the smoke removal rate among the three smoke control systems was small when the opening was closed. The pressurization system showed less smoke removal rate than the other two systems when the opening existed, and hence is not recommended for subway stations with large openings. It was also shown that a less flowrate in the ventilation system leads to a much longer smoke removal time.