微型电脑应用2025，Vol.41Issue(9)：146-149,153,5.
基于BIM+CNN+LSTM的地铁施工安全风险预警平台研究
Research on Subway Construction Safety Risk Warning Platform Based on BIM+CNN+LSTM
李万博 1马伟伟 2郭永亮 2杨杭2
作者信息
- 1. 佛山市地铁建设有限公司,广东,佛山 528000
- 2. 中铁三局集团有限公司,山西,太原 030000
- 折叠
摘要
Abstract
In order to effectively reduce occurrence of safety accidents during subway construction,a subway construction safety risk warning platform is constructed based on building information modeling(BIM),visualization,Internet of Things,sensors and other technologies.The platform adopts C/S architecture,and can achieve various functions such as user management,sensor management,digital-modular correlation,data viewing,and warning management.At the same time,the platform uses the convolutional neural network-long short-term memory(CNN-LSTM)model to extract features and predict time series from monitoring data with multiple parameters,high frequency,and huge information content.Taking the Foshan urban rail transit line 3 project as an example,the built platform is applied to the monitoring of subway construction settlement and deformation,and fourteen types of monitoring data are analyzed,with the maximum settlement and deformation as the output target.The mean square error of the platform prediction is only 4.2％,which is 4.7％lower than the BP neural network(BPNN)model and 1.3％lower than the LSTM model.The regression determination coefficient R2＞0.95.The platform has good generaliza-tion ability and warning effect.关键词
地铁盾构施工/安全风险/预警平台/BIMKey words
subway shield tunneling construction/safety risk/warning platform/BIM分类
建筑与水利引用本文复制引用
李万博,马伟伟,郭永亮,杨杭..基于BIM+CNN+LSTM的地铁施工安全风险预警平台研究[J].微型电脑应用,2025,41(9):146-149,153,5.
