The MD-Yolov7-enabled Smart Detection Approach for Substation Environmental Targets

Due to the difference of live voltage of different live equipment in substation, the safe working distance is different, so it is very important to correctly detect the substation equipment in the process of autonomous operation. How we can accurately identify charged devices. We use deep learning to fit Gaussian stretch position information to extend lightweight interference data, and introduce MD- Yolov7 environmental target intelligent detection method. In addition, we built an experimental test platform and carried out the verification experiment. The experimental confirmed that the deep learning approach achieves 98.7% average accuracy (mAP), and the accuracy is significantly improved. This result not only proves the feasibility of our method, but also shows its accuracy. The method proposed in this paper has a certain reference significance for substation autonomous safety operation detection.