Fault Diagnosis and Improvement of Substation Grounding Network in Substation Engineering Based on Video Image Recognition

With the development of power grid digitization and the explosive growth of current power demand, video image recognition algorithms based on deep learning can effectively cope with the increasingly complex grounding grid fault location system. In the improvement of the entire system model, how to improve the accuracy of line selection and the stability of the device is currently a key issue that needs to be urgently solved. This article analyzes the influencing factors of grounding grid design and fault diagnosis through examples of grounding grid design problems, applies deep learning based image recognition algorithm, and combines convolutional neural network based grounding grid fault process to draw the following conclusion: the fault area location model improves the accuracy of line selection by 8.64% compared to traditional methods under five different branches, and the improved average resistance in terms of device stability is 0.296 Ω, indicating that its power device is more stable. This study aims to establish a good electrical connection between substation equipment and personnel and the ground to ensure personal safety and normal operation of the equipment.