数字孪生驱动的潮汐高差区域跨海大桥钢围堰安全监测方法  

Safety Monitoring Method of Steel Cofferdams Driven by Digital Twin for Sea-Crossing Bridges in Tidal Height Difference Regions

作　　者：刘彬 LIU Bin(The 5th Engineering Co.,Ltd.of China Railway 16th Bureau Group,Tangshan Hebei 064000,China)

机构地区：[1]中铁十六局集团第五工程有限公司,河北唐山064000

出　　处：《铁道建筑技术》2025年第1期116-121,共6页Railway Construction Technology

基　　金：中铁十六局集团有限公司科技研发计划项目(19-16C)。

摘　　要：钢围堰过程位移监测是钢围堰使用过程中最为重要的决策指标,但基于人为全站仪测量的传统位移监测方式存在主观性强、时效性差的缺陷,且潮汐高差区域的潮汐变化大,对大桥以及围堰的安全带来了巨大的挑战。随着数字孪生技术的持续深入,基于传感器的钢围堰数字监测技术不仅能够解决人为监测存在的不足,且能保证其准确性及可靠性。本文基于视觉几何网络(Visual Geometry Group,VGG)提出了一种改进局部位移监测网络,该方法基于有限元模拟数据集,通过将实时钢围堰局部监测状态转换为RGB特征数据输入VGG网络,据此训练出时间段内最贴近实际情况的有限元状态监测模型。所提出的方法不仅解决了既有监测方案存在的缺陷,同时提供了一种更为高效、可靠的工程监测方案,这也为基于数字孪生技术的安全监测方法带来了更高效、可靠的解决方案。The process displacement monitoring of steel cofferdam is the most important decision-making index in the use of steel cofferdam,but the traditional displacement monitoring based on man-made total station measurement has the defects of strong subjectivity and poor timeliness,and the large tidal changes in the tidal height difference area bring great challenges to the safety of the bridge and cofferdam.With the continuous deepening of digital twin technology,the sensor-based digital monitoring technology of steel cofferdam can not only solve the deficiencies of human monitoring,but also ensure its accuracy and reliability.The article proposes an improved local displacement monitoring network by Visual Geometry Group(VGG),which is based on the finite element simulation dataset,and by converting the real-time local monitoring state of steel cofferdams into RGB feature data and inputting them into the VGG network,a finite element state monitoring model that is the closest to the actual situation in the time period is trained accordingly.The proposed method not only solves the defects of the existing monitoring scheme,but also provides a more efficient and reliable engineering monitoring scheme,which also brings a more efficient and reliable solution for the safety monitoring method based on digital twin technology.

关 键 词：钢围堰 数字孪生 深度学习 VGG网络 

分 类 号：U445.556[建筑科学—桥梁与隧道工程]