A three-dimensional reconstruction method of the converter transformer based on point cloud

In order to solve the problems of poor accuracy and low efficiency of manual modeling of converter transformer, a set of 3D reconstruction process based on point cloud equipment was proposed. It is verified that the completeness of the data collected by laser scanning technology for converter station equipment point clouds can reach 99.2% with a 50% sampling overlap rate. The RANSAC point cloud coarse registration process is introduced to solve the problem that the initial position of the cloud affects the efficiency of traditional ICP algorithm. The Greedy projection triangulation algorithm is used to reconstruct the point cloud data in 3D. The reconstruction process is verified with the example of the converter transformer point cloud data. In summary, the 3D reconstruction process of converter transformer proposed in this paper can effectively solve the problem of low modeling efficiency of virtual simulation system, and has certain reference significance and promotion value in the construction of virtual simulation system related to the power industry.