Fault Detection of insulator in distribution network Based on YOLOv5s Neural Network

Abstract

In view of the complex background of the current distribution network insulator inspection image, the detection target is small, the defect forms are various, and it is easy to be blocked by equipment or shadows, resulting in false detection and missed detection, and the detection accuracy is low. A grading detection method is proposed. First, the YOLOv5s network is used to locate the insulator area, and on this basis, the DenseNet201 network is used to further distinguish whether there is a fault in the insulator area. The experimental results show that compared with the original YOLOv5s network, the YOLOv5s-based distribution network insulator defect classification detection method can better identify faulty insulators with insufficient feature expression ability under occlusion, and eliminates false detection of background. It can effectively realize the identification and defect detection of insulators in the inspection images of distribution lines.