摘要
针对当前电力设计行业中存在轻量化三维模型设计校验方法自动化水平较低的问题,提出了一种计及电网工程全寿命周期的轻量化三维模型自动化校验方法.该方法基于BIM技术和GIM技术构建电网工程的三维模型,并引入了改进的SLAM技术完成工程的点云特征拼接重建,在此基础上利用改进深度学习模型KPConv完成三维模型节点的点云语义分割比对,从而实现模型的自动化校验.仿真实验结果表明,采用所提方法在相关模型的自动化校验中取得了较高的精度,且在校验误差方面具有较优的稳定性,因此其在电网工程轻量化三维模型的设计领域具有良好的应用前景.
To address the problem of lower automation level of lightweight 3D model design calibration method in the current power design industry,a lightweight 3D model automation calibration method considering the whole life cycle of power grid project was proposed.The as-proposed method was based on BIM and GIM technologies to construct the 3D model of power grid project,and the improved SLAM technology was introduced to complete the reconstruction of point cloud feature stitching of the project,in terms of which the improved deep learning model KPConv was used to complete the point cloud semantic segmentation of 3D model nodes,so as to realize the automatic calibration of the model.The simulation results show that the as-proposed method achieves higher accuracy in the automatic calibration of relevant models and has better stability in respect of calibration errors,so it has good potential for application in the field of designing lightweight 3D models for power grid engineering.
作者
张爱军
白英
邵雪瑾
王新新
ZHANG Aijun;BAI Ying;SHAO Xuejin;WANG Xinxin(College of Energy and Electrical Engineering,Hohai University,Nanjing 210098,China;Substation Design Center,Ningxia Ningdian Electric Power Design Co.Ltd.,Yinchuan 750000,China)
出处
《沈阳工业大学学报》
CAS
北大核心
2023年第5期503-509,共7页
Journal of Shenyang University of Technology
基金
宁夏回族自治区重点基金项目(SGNXJS00JYGC2100046).
关键词
电网工程全寿命周期
BIM技术
SLAM技术
GIM模型
深度学习模型
点云语义分割
自动化校验
三维模型
whole life cycle of power grid project
BIM technology
SLAM technology
GIM model
deep learning model
point cloud semantic segmentation
automatic calibration
three-dimensional model