无人机热红外遥感技术对矿区煤火探测识别研究

Research on coal fire detection and identification in mining areas based on UAV thermal infrared remote sensing technology

提出了1套高效的卫星和无人机遥感煤火快速探测方法,以陕蒙交界煤炭资源开发强度最大的国能神东煤炭集团有限责任公司石圪台煤矿为研究区,采用卫星热红外大范围圈定煤火靶区,通过无人机高分辨率热红外和光学遥感精细探测煤火的具体位置和温度,并分析其存在的问题和提出改进措施。研究发现,利用卫星热红外遥感反演地表温度来进行煤火识别的准确性不够,因卫星数据分辨率较低,各种因素影响导致温度计算误差较大,反演温度普遍偏低,但能够将卫星反演数据作为研究靶区圈定的依据;无人机热红外探测地表分辨率达厘米级,可以准确探测煤火发生的具体位置、范围以及温度,获得了煤火燃烧顶点温度326.2℃,超过300℃的有9处,其他高温火区也符合矿区采掘和煤火发育规律;卫星和无人机结合的煤火探测模式充分发挥了各自的优势,能够做到高效、便捷、准确地识别煤火信息,结合多源数据的综合解译能够为小尺度矿区煤火探测形成一套完整的操作技术流程。

In this paper,an efficient method of rapid coal fire detection based on satellite and UAV remote sensing is proposed.The method takes Shigetai Coal Mine of China Energy Shendong Coal Group Co.,Ltd.with the highest intensity of coal resource development at the border of Shaanxi and Inner Mongolia,as the research area,uses satellite thermal infrared to circle the coal fire target area,finely detects the specific location and temperature of the coal fire by unmanned aerial vehicle(UAV)high-resolution thermal infrared and optical remote sensing,and analyzes the existing problems and proposes improvement measures.The research shows that the accuracy of satellite thermal infrared remote sensing inversion of surface temperature for coal fire identification is not enough,because of the low resolution of satellite data and the influence of various factors leading to large temperature calculation errors,and the inversion temperature is generally low,but the satellite inversion data can be used as the basis for research target area delineation;the surface detection resolution of UAV thermal infrared reaches centimeter level,which can accurately detect the specific location,range and temperature of coal fire occurrence;in this study,a coal fire burning apex temperature of 326.2°C is obtained,and there are nine locations exceeding 300°C,and the other high temperature fire areas are also consistent with the mining and coal fire development rules.The combination of satellite and UAV coal fire detection mode gives full play to their respective advantages and can achieve efficient,convenient and accurate identification of coal fire information,combined with the comprehensive interpretation of multi-source data,it can form a complete set of operational technology processes for coal fire detection in small-scale mining area.