Terahertz Tomography of Substation Grounding Grid Fault based on Improved Generative Adversarial Networks

Due to the difficulty in replacing the grounding grid of urban integrated substations, any grounding grid failure can have a significant impact on the safety of the substation and surrounding buildings. This paper studies the terahertz tomography detection technology based on an improved generative adversarial network for fault detection of substation grounding grids. According to the characteristics of low resolution and high noise of terahertz image, uses terahertz tomographic image detection technology based on improved generative adversarial network. Firstly, the improved generation countermeasure network to improve the detail processing ability of the image; Secondly, uses the layer hopping connection method to effectively improve the context information reference ability of the image, so as to make the image more detailed and more specific. Finally, the method proposed in this paper is compared with other traditional processing methods, this method is more suitable for substation grounding grid fault detection.