摘要
绝缘子的损坏在高压输电线路里是最常见的一种电力故障,其缺陷检测是巡检的重要环节。大部分现有的绝缘子缺陷检测算法通常都是基于YOLOv5模型,然而现有的基于YOLOv5模型的算法检测精度不高,并且无法满足无人机实时检测的需求。因此,提出了一种基于坐标注意力机制的轻量型目标检测算法。实验结果表明改进后的模型平均准确度能达到96.2%,召回率95.1%,其精度和实时性都优于传统的YOLOv5算法和其他绝缘子检测算法。
Insulator damages are the most frequent type of power faults in high-voltage transmission lines,the defect detection of which is an important part of inspection.Most available insulator defect detection methods are usually based on the YOLOv5 model.However,those methods have low detection accuracy and do not meet the needs of real-time UAV inspection.In order to solve these problems,this paper proposes a lightweight detection algorithm based on attention mechanisms.The detection results show that the proposed algorithm can achieve an average accuracy of 96.2%and a recall rate of 95.1%.In terms of accuracy and real-time performance,the algorithm proposed in this paper outperforms the traditional YOLOv5 algorithm and other insulator detection algorithms.
出处
《工业控制计算机》
2023年第7期79-81,共3页
Industrial Control Computer
