南京信息工程大学学报2026，Vol.18Issue(2)：278-288,11.DOI:10.13878/j.cnki.jnuist.20250402003
面向变电站数字孪生的三维对象匹配方法研究
A novel 3D object matching method for substation digital twins
摘要
Abstract
Significant discrepancies exist between as-built point clouds(from construction/operation)and as-de-signed BIM equipment 3D models(from design phase)in substation digital twins.These differences lead to incon-sistency in the quantity,location,and distribution of power equipment in different scenarios,further hindering the ef-fective application of BIM model in the digital twin context.Here,we propose a novel 3D object matching method based on a first-order distance-topology similarity index and adaptive spatial position correction.First,a first-order distance-topology compatibility index is constructed to better represent the location features of 3D objects;second,the similarity between 3D objects is calculated to identify and extract both precise and candidate matching pairs;third,a set of line-of-sight constraint equations is constructed to enable iterative self-correction of the spatial position for BIM models in the candidate matching pairs.We evaluated our method on 3D datasets from two complex substa-tions,and the overall accuracy reaches 97.92%for the first substation and 98.54%for the second.Experimental re-sults demonstrate that our method significantly outperforms state-of-the-art alternatives in accuracy and reliability for 3D object matching in complex substation digital twin scenarios.关键词
实景点云对象/BIM三维模型/三维对象匹配/一阶距离-拓扑兼容性指标/位置迭代纠正Key words
real-scene point cloud objects/BIM 3D models/3D object matching/first-order distance-topology in-dex/iterative position correction分类
信息技术与安全科学引用本文复制引用
王兴华,董晗拓,蒙彦锡,王流火,王彦峰,贾涛..面向变电站数字孪生的三维对象匹配方法研究[J].南京信息工程大学学报,2026,18(2):278-288,11.基金项目
广东电网有限责任公司科技项目(030000KC23040114) （030000KC23040114）
