基于改进YOLOv7的输电铁塔塔基检测算法

输电塔作为整个电力传输系统最重要的组成部分之一,需要及时对输电塔进行检测保证塔基的稳固以保障后期的使用。针对无人机采集到的输电塔图像存在背景复杂、背景与目标塔基对比度低、小目标及塔基不完整等问题,提出了基于改进YOLOv7的输电塔塔基检测算法。首先,通过无人机采集不同地形地貌的输电塔图像,构建高质量数据集。然后,在原始YOLOv7的Backbone层中加入卷积注意力模块CBAM注意力机制,以提高输电塔塔基特征的提取能力。最后,引入WIoU v3代替原坐标损失函数CIoU,以提高目标检测任务的准确性和稳定性。在该数据集上,使用改进后的YOLOv7算法与目前主流的目标检测算法进行对比实验,实验结果中所提算法的mAP值高达99.93%,比原始YOLOv7提高2.19%,FPS值为37.125,满足实时检测需求,算法的整体性能较好。实验验证了所提算法在塔基检测上的可行性和有效性,为后续塔基区周围水土情况的研究奠定了基础。

Study on Improvement of Seismic Performance of Transmission Tower Using Viscous Damper

The earthquake resistance of transmission tower has been often discussed from the viewpoint of reinforcing the foundation of steel tower, but there are also few studies considering the damping characteristics of the tower. This paper focuses on the viscous damper which has been adopted for seismic reinforcement of bridges in recent years. The purpose of this study is to improve the seismic performance of steel tower by giving the high damping to the tower. We construct a single tower model considering the influence of transmission line, and then simulate the vibration characteristics and seismic behavior of the tower by the eigenvalue analysis and the dynamic response analysis. The results show that the transmission tower with viscous damper can reduce its own response effectively and drastically. This research concludes that it is necessary to consider the extreme increase of steel tower＇s response depending on the seismic wave and the collapse of steel tower can be avoided by using the optimum damper in the design of the transmission tower.