基于改进目标检测网络的输电线路舞动检测OA
Transmission Line Galloping Detection Based on Improved Object Detection Network
为检测输电线路舞动情况,保障输电网络运行稳定性与可靠性,提出一种基于改进目标检测网络的输电线路舞动检测方法,对输电线路的舞动进行检测.首先,使用目标检测网络YOLO11 的旋转检测框对监控视频中的输电线路进行检测,使得检测框能够更好地包围目标位置;其次,利用融合了BiFormer注意力机制的C3 模块替换一部分C3k2 模块,利用简化的快速空间金字塔池化模块替换快速空间金字塔池化模块并引入PIoU损失函数,对网络进行改进,提高网络对图像的特征提取能力以及检测精度;最后,对检测框所检测到的输电线路进行位置提取并绘制舞动轨迹.实验结果表明,与未改进的网络相比,本文方法具有更好的舞动检测效果及检测精度,其精确率及均值平均精度分别提高了 2.4%、1.3%.
To detect the phenomenon of transmission line galloping and ensure the stability and reliability of transmis⁃sion network,a transmission line galloping detection method based on an improved object detection network is proposed to detect transmission line galloping.First,the transmission line in the monitoring video is detected by oriented bound⁃ing boxes of the object detection network YOLO11,so that the bounding boxes can better surround the object position.Second,one part of the C3k2 module is replaced by the C3 module which integrates a BiFormer attention mechanism,a simplified spatial pyramid pooling-fast(SPPF)module is used to replace the SPPF module,and a PIoU loss function is introduced to improve the network and enhance its image feature extraction capability and detection accuracy.Finally,the position of the transmission line detected by the bounding boxes is extracted,and the galloping trajectory is drawn.Experimental results show that compared with those of the unimproved network,the precision rate and mean average ac⁃curacy of the improved network are improved by 2.4%and 1.3%,respectively,indicating that the proposed method has better galloping detection effect and detection accuracy.
付文龙;熊浩伟;陈慧;贾骏;文斌;杨楠
三峡大学电气与新能源学院,宜昌 443002||三峡大学梯级水电站运行与控制湖北省重点实验室,宜昌 443002三峡大学电气与新能源学院,宜昌 443002中国南方电网有限责任公司超高压输电公司曲靖局,曲靖 655000国网江苏省电力公司电力科学研究院,南京 211103三峡大学电气与新能源学院,宜昌 443002三峡大学电气与新能源学院,宜昌 443002
信息技术与安全科学
输电线路改进的目标检测网络旋转目标检测舞动检测
transmission lineimproved object detection networkrotating object detectiongalloping detection
《电力系统及其自动化学报》 2026 (3)
99-105,7
湖北省自然科学基金联合基金资助项目(2024AFD409).
评论