Coal refuse pile fires are a serious health, safety, and environment hazard. The fires in Guoshan Coal Refuse Pile affect daily life of the local residents. Although conventional fire control methods (water injection, grouting and surface sealing) were used to control the fires, they rekindled soon after the fire-fighting engineering was quitted. Thus, effective methods need to be applied to extinguish the fires completely without reignition. In this paper, the environmental impact of the fire on the local land, water and air is illustrated by sample analysis and photos. The self-potential method, which achieves remote detection from the ground surface, is adopted to delineate the scope of the fires to give support for the efficient and economical extinguishment of them. The three-phase foam containing water, gas and loess is proposed to control and extinguish the fires. The corresponding fire-fighting system is also invented to fit the field work and put out the fires more efficiently and conveniently. The temperature monitoring result shows that the fires are extinguished successfully with no sign of reignition.
23 May 2022 in IEEE Transactions on Geoscience and Remote Sensing
Coal fires have led to severe ecological disasters worldwide and threatened the health and safety of nearby residents to a certain extent. It is critical to detect the scope and extent of coal fires in order to control and extinguish them. In this article, the unmanned aerial vehicle (UAV), thermal infrared imaging, and oblique photogrammetry technologies were integrated to create a 3-D land surface temperature model of the Chunjingwa Coal Fire in Shanxi Province, China. Hundreds of visible and thermal infrared images were collected from five flight survey lines. The thermal infrared images were preprocessed in batches to meet the requirements of 3-D modeling. The construction of the thermal infrared 3-D model of the fires was achieved successfully, allowing for the visualization of the surface temperature field in true 3-D. Four high-temperature zones were delineated according to the 3-D land surface temperature model. Furthermore, the features of ground objects were clearly depicted by the visible 3-D model constructed by visible images, and the digital elevation model (DEM) was established based on the spatial coordinate contained in the visible 3-D model. The research results provide technical support for the subsequent firefighting plan and a reference for the high-precision detection of other coal fires around the world.
18 April 2024 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Coal gangue is the byproduct of coal mining and preparation. Due to its susceptibility to spontaneous combustion, it causes significant environmental hazards when piled up in open dumps. The fire not only harms the environment but also poses health risks to nearby residents. Therefore, accurately identifying the location and extent of the coal gangue dump fire is crucial for effective fire suppression. However, the clarity and precision of detection face challenges. This article proposes a novel technique that utilizes thermal infrared oblique photography captured by unmanned aerial vehicle to create three-dimensional (3-D) surface temperature field models of fire areas. In addition, batch preprocessing software is designed to standardize the temperature threshold and color distribution across each thermal infrared image, and the effects of different temperature thresholds are investigated. The article focuses on the spontaneous combustion area of the south coal gangue dump at the Antaibao open-pit mine in Shanxi Province. The findings demonstrate the rapid and precise reconstruction of the 3-D temperature field in the fire area, the elimination of pseudo-high-temperature areas, and the accurate identification of the fire within the coal gangue dump. This method facilitates effective fire suppression and controls measures for the burning coal gangue dump, reducing the associated risks.
Computer vision technology has broad application prospects in the field of intelligent fire detection, which has the benefits of accuracy, timeliness, visibility, adjustability, and multi-scene adaptability. Traditional computer vision algorithm flaws include erroneous detection, detection gaps, poor precision, and slow detection speed. In this paper, the efficient and lightweight YOLOv5s model is used to detect the fire flame and smoke. The attention mechanism is embedded into the C3 module to enhance the backbone network and maximize the algorithm's suppression of invalid feature data. Alpha CIOU is adopted to improve the positioning function and detection target. At the same time, the concept of transfer learning is used to realize semi-automatic data annotation, which reduces training expenses in terms of manpower and time. The comparative experiments of 6 distinct fire detection algorithms (YOLOv5 and 5 optimization algorithms) are carried out. The results indicate that the self-attention mechanism based on the Transformer structure has a substantial impact on enhancing target detection precision. The improved location function based on Alpha CIOU aids in enhancing the detection recall rate. The average recall rate of fire detection of the YOlOv5+TR+αCIOU algorithm is the highest, which is 68.5%, clearly outperforming other algorithms. Based on the surveillance video, this optimization algorithm is utilized to detect a fire in a factory, and the fire is detected in the 9th second when it starts to appear. The results demonstrate the algorithm's viability for real-time fire detection.