基于BIM技术的配网数字化平台设计与实现  

Design and Implementation of Distribution Network Digital Platform Based on BIM Technology

作　　者：许志华 赵俊生[2] XU Zhihua;ZHAO Junsheng(China Southern Power Grid Digital Grid Research Institute Co.,Ltd.,Guangzhou 510630,China;Beijing North-Star Technology Development Co.,Ltd.,Beijing 100010,China)

机构地区：[1]南方电网数字电网研究院有限公司,广东广州510630 [2]北京洛斯达科技发展有限公司,北京100010

出　　处：《微型电脑应用》2023年第7期73-77,共5页Microcomputer Applications

基　　金：南网数研院科技项目(670000HK42200003)。

摘　　要：为了进一步提升配电网建设效率,提出一种基于BIM技术的配网数字化平台设计。平台采用具有数据层、服务层与应用层的BIM架构,通过云技术和NoSQL数据库,实现BIM海量数据管理与储存。通过时间序列检测出异常数据,建立小波基神经网络,利用梯度学习方式训练网络,将动量项引入数据清洗过程中,满足数据序列流速快的要求,实现配网数字化有效显示。仿真实验结果表明用所提出的方法构建的平台数据储存极值较高、平台稳定性更高、设计耗时更短、平台应用效果更好,能够提高配电网建设效率,提高配电网运行的稳定性。In order to further improve the efficiency of distribution network construction,a design scheme of distribution network digitization platform based on BIM technology is proposed.The platform establishes BIM architecture including data layer,service layer and application layer,and realizes the management and storage of BIM massive data by cloud technology and NoSQL database.Using time series to detect abnormal data,a wavelet-based neural network is established,and the network is trained by gradient learning.Momentum term is introduced into the process of data cleaning,which meets the requirement of fast flow rate of data series,and realizes the digitalization and effective display of distribution network.The simulation results show that the platform built by this method has higher limit data storage,stronger platform stability,shorter design time and better application effect,and can improve the efficiency of distribution network construction and the stability of distribution network operation.

关 键 词：建筑信息模型 配网数字化平台 云储存 异常检测 神经网络 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]