A 2.5-D Substation Terrestrial Laser Scanning Network Planning Algorithm Based on Ray Weighting

Terrestrial laser scanning (TLS) network planning plays a crucial role in efficiently and accurately capturing point clouds of substation equipment. However, due to the diversity of equipment types and significant occlusions in substation environments, along with the need to position scans on roads for construction safety, obtaining high-quality point clouds with fewer scanners remains a challenge. To address this issue, we propose a 2.5-D TLS scanning network planning method for substations on the basis of ray weighting. First, we sample the substation scene model via triangular mesh. Next, we perform vertical slicing of the resulting point clouds. Then, we extract the bounding boxes of the sliced point clouds in the XY plane and stack them to generate the 2.5-D models. Visibility ray weights are calculated on the basis of propagation distance and equipment size, leading to the construction of a global topological visibility map under the constraint of the incidence angle. This map drives the network layout by representing changes in information gain. By applying this method, TLS networks are generated for two substation model datasets, followed by accuracy validation. The results demonstrate that, compared with other methods, the proposed approach reduces the number of scan setups by 3–5, increases by 7.4% –8.9% in the coverage, and improves the proportion of point clouds meeting the quality standard by 8.8% –9.1%. This method effectively captures high-quality point clouds in substation scenes and holds significant potential for TLS network planning, providing essential methods and data support for substation scene scanning.