Flame Smoke Research Articles

Application of a real-time flame smoke detection algorithm based on improved YOLOv7

Flame and smoke detection is a critical issue that has been widely used in various unmanned security monitoring scenarios. However, existing flame smoke detection methods suffer from low accuracy and slow speed, and these problems reduce the efficiency of real-time detection. To solve the above problems, we propose an improved YOLOv7(You Only Look Once) algorithm for flame smoke mobile detection. The algorithm uses the Kmeans algorithm to cluster the prior frames in the dataset and uses a lightweight CNeB(ConvNext Block) module to replace part of the traditional ELAN module to accelerate the detection speed while ensuring high accuracy. In addition, we propose an improved CIoU loss function to further enhance the detection effect. The experimental results show that, compared with the original algorithm, our algorithm improves the accuracy by 4.5% and the speed by 39.87%. This indicates that our algorithm meets the real-time monitoring requirements and can be practically applied to field detection on mobile edge computing devices.

Comparative study on combustion and emission characteristics of gas-to-liquid with diesel on a burner and a naturally aspirated compression ignition engine

ABSTRACT Due to the consumption of fossil fuels and the generation of harmful emissions by diesel engines, researchers are seeking alternative fuels for diesel, for example, gas-to-liquid (GTL). However, most studies on combustion and emission of GTL focused on the fuel burned on turbocharged heavy-duty diesel engines. The purpose of this study is to reveal the combustion, fuel economy, and emission characteristics of GTL by a burner and a naturally aspirated (NA), fuel direct injection compression ignition light-duty engine. Comparative experiments of the GTL to fossil diesel were carried out, respectively. The burner has good consistency with the diesel engine in revealing emission trends. The results of the burner test showed that GTL produced less flame smoke and its flame height was 5.3% lower than diesel. The results of the engine bench test revealed that the combustion and emissions of the GTL with diesel were significantly different. Compared to diesel, the ignition of GTL began earlier from 1.5°CA to2.5°CA though the combustion was slower in contrast to diesel. The diffusion combustion maximum pressure was considerably 4.8% higher and occurred earlier with 2.2°CA-3°CA, and the premixed combustion peak pressure of GTL was slightly beyond the diesel. The peak of premixed combustion heat release rate (HRR) for GTL was less than that of diesel by 18.9%. While the peak HRR of diffusion combustion was greater by 38.4%. The maximum PRR value of GTL was lower than that of diesel leading to soft combustion. In all test conditions, the brake specific fuel consumption of GTL averagely reduced by 4.7% against diesel, and the brake thermal efficiency was on average 4.3% higher. GTL engine can reduce regular emissions of carbon monoxide, hydrocarbons, nitrogen oxide (NOx), smoke, and particle number (PN) simultaneously. It is worth noting that GTL improved the trade-off issue of NOx and smoke, simultaneously reducing 18.8% NOx and 24.1% smoke on average. The experimental results reveal the GTL is an appropriate alternative fuel for light-duty diesel engine without modification of mechanical fuel system.

Numerical simulation of a continuously released ethanol spill fire in an inclined tunnel with longitudinal ventilation

Under the influences of the chimney effect and longitudinal ventilation, the risks of continuous spill fires in inclined tunnels are catastrophic with great complexity and uncertainty. In this paper, the spread and combustion behaviour of a continuous spill fire under the effect of longitudinal ventilation in an inclined tunnel is studied and compared with that in a horizontal tunnel through fire dynamics simulator (FDS) software. The critical time shrink point (SP) to reach the maximum combustion area (MCA), the change rates of the dimensionless combustion area A* and stable combustion area (SCA), the flame propagation speed, and the dimensionless burning rate W* values and heat transfer mechanisms of the combustion fuel layer, tunnel wall and flame smoke are analysed; the prediction equation between the dimensionless burning rate W* and the dimensionless wind speed V* of the upstream and downstream spill fire are identified. The results show that longitudinal ventilation leads to a nonlinear increase in the MCA, and the longitudinal slope leads to a great decrease in the critical time SP. In the stable combustion stage, the SCA in the downstream spill fire is greater than that in the upstream spill fire. Additionally, there are significant differences in the flame propagation phenomenon of spill fires. On the windward side, the maximum flame propagation speed of the upstream spill fire is 3.8 times that of the downstream spill fire; however, on the leeward side, the maximum flame propagation speed of the upstream spill fire is only 18% of that of the downstream spill fire. According to the prediction equation, with the increase in the dimensionless wind speed V*, the dimensionless burning rate W* decreases gradually; the dimensionless burning rate W* of the upstream spill fire is always greater than that of the downstream spill fire. The increase in thermal feedback caused by the inclined flame is the main reason for the significant difference in the spread of the spill fire in the ventilated inclined tunnel, causing the flame propagation speed of the −10% slope on the leeward side to gradually approach that of the −5% slope.

Flame smoke detection algorithm based on YOLOv5 in petrochemical plant

PurposeFire smoke detection in petrochemical plant can prevent fire and ensure production safety and life safety. The purpose of this paper is to solve the problem of missed detection and false detection in flame smoke detection under complex factory background.Design/methodology/approachThis paper presents a flame smoke detection algorithm based on YOLOv5. The target regression loss function (CIoU) is used to improve the missed detection and false detection in target detection and improve the model detection performance. The improved activation function avoids gradient disappearance to maintain high real-time performance of the algorithm. Data enhancement technology is used to enhance the ability of the network to extract features and improve the accuracy of the model for small target detection.FindingsBased on the actual situation of flame smoke, the loss function and activation function of YOLOv5 model are improved. Based on the improved YOLOv5 model, a flame smoke detection algorithm with generalization performance is established. The improved model is compared with SSD and YOLOv4-tiny. The accuracy of the improved YOLOv5 model can reach 99.5%, which achieves a more accurate detection effect on flame smoke. The improved network model is superior to the existing methods in running time and accuracy.Originality/valueAiming at the actual particularity of flame smoke detection, an improved flame smoke detection network model based on YOLOv5 is established. The purpose of optimizing the model is achieved by improving the loss function, and the activation function with stronger nonlinear ability is combined to avoid over-fitting of the network. This method is helpful to improve the problems of missed detection and false detection in flame smoke detection and can be further extended to pedestrian target detection and vehicle running recognition.

Flame and Smoke Detection Algorithm Based on ODConvBS-YOLOv5s

Real-time and accurate detection of flame and smoke is an important prerequisite to reduce the loss caused by fire. There exists some problems in traditional flame and smoke detection algorithm, such as low accuracy, high miss rate, low detection efficiency and low detection rate of small targets. This paper proposes a YOLOv5s flame smoke detection algorithm based on ODConvBS. Firstly, in the YOLOv5s backbone network, the ordinary convolution block is replaced by ODConvBS to realize the extraction of attention features of the convolution kernel; Secondly, Gnconv is introduced into Neck to improve the model’s high-order spatial information extraction ability; then the Shuffle Attention module is added at the end of Neck to facilitate the fusion of different groups of features; At last, in the prediction section, the SIOU loss function, which can account for the angle of the prediction frame vector, is utilized to speed up model convergence. When utilizing the self-made flame and smoke data set, the upgraded YOLOv5s model mAP grew by 9.3%.At the same time, the accuracy rate and the recall rate and the detection speed increased to 83.5%, 83.7%, 33.3FPS respectively.

Laboratory Flame Smoke Detection Based on an Improved YOLOX Algorithm

Fires in university laboratories often lead to serious casualties and property damage, and traditional sensor-based fire detection techniques suffer from fire warning delays. Current deep learning algorithms based on convolutional neural networks have the advantages of high accuracy, low cost, and high speeds in processing image-based data, but their ability to process the relationship between visual elements and objects is inferior to Transformer. Therefore, this paper proposes an improved YOLOX target detection algorithm combining Swin Transformer architecture, the CBAM attention mechanism, and a Slim Neck structure applied to flame smoke detection in laboratory fires. The experimental results verify that the improved YOLOX algorithm has higher detection accuracy and more accurate position recognition for flame smoke in complex situations, with APs of 92.78% and 92.46% for flame and smoke, respectively, and an mAP value of 92.26%, compared with the original YOLOX algorithm, SSD, Faster R-CNN, YOLOv4, and YOLOv5. The detection accuracy is improved, which proves the effectiveness and superiority of this improved YOLOX target detection algorithm in fire detection.

基于卷积神经网络室内火焰烟雾识别

基于卷积神经网络室内火焰烟雾识别

The Path Search Be Applied Algorithm for Spread of Flame and Smoke in Fire Simulation

Fire is one of the disasters that often occur in everyday life. There has been much research done to build a fire Simulator. In this paper, we propose a fire Simulator that is applied optimal path search algorithm to arrange the quick evacuation. This simulator considers the structure of a building and the location where the fire has initially occurred, applying it to the simulator. Especially when implementing the training simulation for firefighters, optimized evacuation path for the required virtual rescuer awaiting rescue was informed to implement more realistic firefighting training simulator. For the simulation for the test, fixed data was set to apply the shortest path search algorithm and flame smoke spread algorithm. As a result of the test, the optimized path was derived through avoiding the flame occurred from flame smoke spread algorithm and following the shortest path search algorithm. Keyword: Fire Simulator, optimal path, Greedy method, Binary programming

Radiation from Buoyant Turbulent Diffusion Flames

Abstract The relationship between the radiant fraction, X R, of the total heat release rate from buoyant turbulent diffusion flames and a fuel's laminar flame smoke point is refined and extended to include: additional hydrocarbon fuels, fuel dilution with nitrogen and a range of oxygen/nitrogen ambient environments. Correlation of the data allows one to predict X R in terms of the: (1) fuel smoke-point laminar flame height, l 3, (or corresponding heat release rate Q SP ), (2) adiabatic stoichiometric flame temperature, T ad of the supplied reactants, and (3) stoichiometric oxidant to fuel mixture mass ratio S. The existence of such a general correlation for S ≥ 12 suggests that the buoyant flame radiation is effectively controlled by only these three reactant properties. The present correlation does not apply for S < 12 because of the increased importance of gaseous radiation for reduced 5. The above correlation of turbulent flame ;XR with the laminar-flame smoke point is achieved by consistently referring both (turbulent and laminar) flame systems to “Standard Conditions” (S = 15 and T ad = 2200 K.). Fuel mixture mass concentrations ranged from 0.139 to 1.00; while oxidant oxygen mass concentrations ranged from 0.ISS to 0.302.

Who do Diffusion flames Emit smoke

The causes of smoke emission from free gaseous laminar diffusion flanles are investi-gated. Soot concentration, temperature and velocity profiles were measured around the smokepoint for a range of fuels. Smoke was emitted when the temperature in the burnout region droppedbelow about 1300 K.A smoking flame stopped emitting smoke when the temperature in the burnout region wasslightly raised locally by absorption of energy from a COz laser beam.The temperature drop in the flame is strongly coupled to soot concentration through radiationlosses. Soot formation rates are in turn influenced by heat losses and residence times near thefuel nozzle tip. Thus conditions near the nozzle influence the smoke point. The results are alsodiscussed in context with turbulent diffusion flame smoke emission.