To address the issue of insufficient panoramic information perception capability of insulation arms in large-scale distribution network operations, a method based on AI and an improved tunicate swarm algorithm is proposed for panoramic information perception of distribution network operation insulation arms. AR equipment and panoramic cameras are employed to monitor panoramic information during distribution network operations. Fusion terminals are used to integrate the multi-source and heterogeneous data collected by the panoramic cameras and AR devices. By introducing a Poisson reconstruction mechanism, the fused monitoring information is converted into continuous point cloud surfaces, and a panoramic environmental spatial database is constructed. An AI-based image ICP registration algorithm is utilized to accomplish 3D point cloud panoramic imaging modeling of the insulation arm in distribution network operation. The particle swarm optimization algorithm is incorporated to enhance the tunicate swarm algorithm for optimizing the modeled point cloud, thereby achieving panoramic information perception for insulation arms in distribution network operations. Experimental results demonstrate that the curvature error of this method for spatial data labeling of power equipment, insulation arms, and obstacles is below the required threshold. The similarity between objects in the constructed 3D point cloud model of the distribution network operation insulation arm and the actual scene exceeds 90%, and the RE value for panoramic perception is high, indicating significantly improved perception accuracy.
Wan et al. (Mon,) studied this question.