This paper presents an experimental study of a prestressed lightweight concrete platform model with a tank and for five steel-columns. This platform can be used not only for extraction but also for storage of oil and ...This paper presents an experimental study of a prestressed lightweight concrete platform model with a tank and for five steel-columns. This platform can be used not only for extraction but also for storage of oil and is suitable for the Bohai Sea and other shallow seas of China. The platform is subjected to temperature. load, or both. The corresponding temperature distribution. strains, cracks. and vulnerable parts of the platform are analyzed respectively. By use of the finite element method and empirical formulas, the temperature field of the model is analyzed. The results agree with the experimental results, thereby verifying! the reliability of these two calculating methods. The paper provides an experimental basis for the des sign of the bearing capacity and normal service state of prestressed concrete platforms.展开更多
无人机载平台中的目标检测在军事和民用领域具有重要的应用价值.然而,现有的检测方法通常侧重于多尺度目标检测,缺乏对小目标的优化,且模型复杂度过高,难以在资源受限的机载平台中应用.为此,本文提出了一种面向无人机载平台的轻量级小...无人机载平台中的目标检测在军事和民用领域具有重要的应用价值.然而,现有的检测方法通常侧重于多尺度目标检测,缺乏对小目标的优化,且模型复杂度过高,难以在资源受限的机载平台中应用.为此,本文提出了一种面向无人机载平台的轻量级小目标检测算法YOLOH(You Only Look One Head).首先,针对小目标对基准网络优化,移除深层特征以减少模型参数量,增加浅层特征以获取小目标信息.其次,在特征融合部分加入NAM注意力,增强对小目标的感知能力.接着,设计了多感受野聚焦模块MRFF,以挖掘特征图的感受野信息,增强模型的多尺度检测能力.最后,使用LAMP算法对模型剪枝,去除冗余神经元以压缩模型.实验结果表明,与YOLOv8s相比,YOLOH的模型参数量和计算量分别减少了92%和35%,FPS提高了57%.在VisDrone2019和CARPK数据集上AP_(S)分别提高了3.3%和3.7%.与其他轻量级模型相比,所提YOLOH具有最佳的整体性能,同时平衡了模型大小、精度和推理速度,为无人机载平台的目标检测提供了有效的解决方案.展开更多
基金The project was financially supported by the National Natural Science Foundation of China(Grant No.59895410)
文摘This paper presents an experimental study of a prestressed lightweight concrete platform model with a tank and for five steel-columns. This platform can be used not only for extraction but also for storage of oil and is suitable for the Bohai Sea and other shallow seas of China. The platform is subjected to temperature. load, or both. The corresponding temperature distribution. strains, cracks. and vulnerable parts of the platform are analyzed respectively. By use of the finite element method and empirical formulas, the temperature field of the model is analyzed. The results agree with the experimental results, thereby verifying! the reliability of these two calculating methods. The paper provides an experimental basis for the des sign of the bearing capacity and normal service state of prestressed concrete platforms.
文摘无人机载平台中的目标检测在军事和民用领域具有重要的应用价值.然而,现有的检测方法通常侧重于多尺度目标检测,缺乏对小目标的优化,且模型复杂度过高,难以在资源受限的机载平台中应用.为此,本文提出了一种面向无人机载平台的轻量级小目标检测算法YOLOH(You Only Look One Head).首先,针对小目标对基准网络优化,移除深层特征以减少模型参数量,增加浅层特征以获取小目标信息.其次,在特征融合部分加入NAM注意力,增强对小目标的感知能力.接着,设计了多感受野聚焦模块MRFF,以挖掘特征图的感受野信息,增强模型的多尺度检测能力.最后,使用LAMP算法对模型剪枝,去除冗余神经元以压缩模型.实验结果表明,与YOLOv8s相比,YOLOH的模型参数量和计算量分别减少了92%和35%,FPS提高了57%.在VisDrone2019和CARPK数据集上AP_(S)分别提高了3.3%和3.7%.与其他轻量级模型相比,所提YOLOH具有最佳的整体性能,同时平衡了模型大小、精度和推理速度,为无人机载平台的目标检测提供了有效的解决方案.
