The Substation Defect Detection Method Based on FEFPN-YOLOv7

In order to ensure the safe and stable operation of substation and enhance the intelligent anomaly monitoring ability of substation, this paper proposes a substation anomaly and equipment defect detection algorithm based on the improved YOLOv7 model. Based on the YOLOv7 algorithm, a novel multi-branch backbone and neck network fusion feature pyramid network (FEFPN) framework is introduced. By adding the information of different channels of backbone network and neck network, the method effectively combines spatial location features with rich semantic information in the image, significantly improving the multi-scale target detection performance and improving the sensitivity to small targets. Experimental tests show that the mAP value of this algorithm is 0.917, which is 8.4% higher than that of YOLOv7. The results show that this method can effectively detect the abnormal and faulty equipment around the substation, and improve the level of intelligent monitoring of the substation.