Visual Modeling of Transmission Tower Air Gap Based on <scp>3D GIS</scp> Technology and Residual Convolution Neural Network

ABSTRACT The condition of the air gap in transmission towers directly impacts transmission safety, and its visualization represents a future research direction. This paper investigates a visualization modeling method for transmission tower air gaps based on 3D GIS technology and residual convolutional neural networks. On a 3D WebGIS platform, this method employs a three‐dimensional visual modeling approach for transmission towers that integrates digital orthophoto image (DOM) and digital elevation model (DEM) data. Using PixelGrid software, DEM and DOM data for transmission towers are automatically acquired, overlaid, and used to display air gap images while obtaining corresponding elevation information, thereby diagnosing the actual air gap status of tower structures. The experimental results show that in terms of modeling error, the proposed method achieves an error of 1.6 ± 0.1 cm under low‐load sunny conditions, demonstrating a significant advantage over other methods. In terms of computational efficiency, the proposed method typically requires less time, with a processing time of 5.8 ± 0.9 s under low‐load sunny conditions, which is notably faster than other methods. In terms of identification stability, its standard deviation is also relatively small, with a value of 0.025 under low‐load sunny conditions, exhibiting the smallest overall fluctuation. This demonstrates that the method proposed in this paper can more accurately, efficiently, and stably identify the air gap status of transmission towers, thereby validating its effectiveness and superiority in identifying air gaps in transmission towers.