It is the basic requirement of the synergetic exploitation of deep mineral resources and geothermal resources to arrange the heat transfer tube in filling body. The heat release performance of filling body directly im...It is the basic requirement of the synergetic exploitation of deep mineral resources and geothermal resources to arrange the heat transfer tube in filling body. The heat release performance of filling body directly impacts on the exploiting efficiency of geothermal energy. Based on heat transfer theory, a three-dimensional unsteady heat transfer model of filling body is established by using FLUENT simulation software. Taking the horizontal U-shaped buried pipe as research object, the variation of temperature field in filling body around buried pipe is analyzed during the heat release process of filling body;the initial temperature of filling body, the diameter of buried pipe, the inlet temperature and inlet velocity of heat transfer fluid influencing of coupling heat transfer, which exists between heat transfer fluid and surrounding filling body within a certain axial distance of buried tube, and influencing of temperature difference between inlet and outlet of heat transfer fluid and on heat transfer performance of filling body are also discussed. It not only lays a theoretical foundation for the synergetic exploitation of mineral resources and geothermal energy in deep mines, but also provides a reference basis for the arrangement of buried pipes in filling body as well as the selection of working conditions for heat transfer fluid.展开更多
在林业管理中,及时发现火灾并识别其规模对于安全防护和治理火灾至关重要。针对现有火灾检测算法存在的精度低、漏检误检和实时性不足等问题,提出一种无人机航拍图像下火灾实时检测算法——MDSYOLOv8。以YOLOv8为基线算法,将骨干网络第...在林业管理中,及时发现火灾并识别其规模对于安全防护和治理火灾至关重要。针对现有火灾检测算法存在的精度低、漏检误检和实时性不足等问题,提出一种无人机航拍图像下火灾实时检测算法——MDSYOLOv8。以YOLOv8为基线算法,将骨干网络第7层卷积模块和颈部网络卷积模块替换成动态蛇形卷积(DSConv),提高算法的特征提取性能,并强化算法对微小特征的学习能力;然后在颈部与检测头之间添加多维协作注意力机制(MCA),加强颈部特征融合,增强算法对小目标的检测能力,并抑制无关背景信息;最后使用SIoU损失函数替换原YOLOv8中的CIoU损失函数,加快算法的收敛速度和回归精度。实验结果表明,MDSYOLOv8在公开数据集KMU上对烟雾目标的检测精度mAP达到95.89%,相较于基线YOLOv8提高了3.33个百分点,具有卓越的检测性能。此外,本研究采集互联网上的无人机航拍火灾图像制作UFF(UAV field fire)数据集,主要对象为火焰和烟雾,包含森林和城市等火灾隐患可能发生场景。在自制数据集UFF上进行深度实验分析,MDSYOLOv8的检测精度达到93.98%,检测速度为54帧/s,并且能同时识别烟雾和火焰两种火灾场景中的主要目标,与主流目标检测方法相比,在检测精度和效率方面均展现出明显优势,更加契合航拍场景下的火灾检测应用。展开更多
基金Projects(51974225,51874229,51674188,51904224,51904225,51504182) supported by the National Natural Science Foundation of ChinaProjects(2018JM5161,2018JQ5183,2015JQ5187) supported by the Natural Science Basic Research Plan of Shaanxi,China
文摘It is the basic requirement of the synergetic exploitation of deep mineral resources and geothermal resources to arrange the heat transfer tube in filling body. The heat release performance of filling body directly impacts on the exploiting efficiency of geothermal energy. Based on heat transfer theory, a three-dimensional unsteady heat transfer model of filling body is established by using FLUENT simulation software. Taking the horizontal U-shaped buried pipe as research object, the variation of temperature field in filling body around buried pipe is analyzed during the heat release process of filling body;the initial temperature of filling body, the diameter of buried pipe, the inlet temperature and inlet velocity of heat transfer fluid influencing of coupling heat transfer, which exists between heat transfer fluid and surrounding filling body within a certain axial distance of buried tube, and influencing of temperature difference between inlet and outlet of heat transfer fluid and on heat transfer performance of filling body are also discussed. It not only lays a theoretical foundation for the synergetic exploitation of mineral resources and geothermal energy in deep mines, but also provides a reference basis for the arrangement of buried pipes in filling body as well as the selection of working conditions for heat transfer fluid.
文摘在林业管理中,及时发现火灾并识别其规模对于安全防护和治理火灾至关重要。针对现有火灾检测算法存在的精度低、漏检误检和实时性不足等问题,提出一种无人机航拍图像下火灾实时检测算法——MDSYOLOv8。以YOLOv8为基线算法,将骨干网络第7层卷积模块和颈部网络卷积模块替换成动态蛇形卷积(DSConv),提高算法的特征提取性能,并强化算法对微小特征的学习能力;然后在颈部与检测头之间添加多维协作注意力机制(MCA),加强颈部特征融合,增强算法对小目标的检测能力,并抑制无关背景信息;最后使用SIoU损失函数替换原YOLOv8中的CIoU损失函数,加快算法的收敛速度和回归精度。实验结果表明,MDSYOLOv8在公开数据集KMU上对烟雾目标的检测精度mAP达到95.89%,相较于基线YOLOv8提高了3.33个百分点,具有卓越的检测性能。此外,本研究采集互联网上的无人机航拍火灾图像制作UFF(UAV field fire)数据集,主要对象为火焰和烟雾,包含森林和城市等火灾隐患可能发生场景。在自制数据集UFF上进行深度实验分析,MDSYOLOv8的检测精度达到93.98%,检测速度为54帧/s,并且能同时识别烟雾和火焰两种火灾场景中的主要目标,与主流目标检测方法相比,在检测精度和效率方面均展现出明显优势,更加契合航拍场景下的火灾检测应用。
