摘要
由于基坑开挖会对地层产生扰动,有可能引起附近建筑物沉陷和偏移,且受自然环境例如持续降水的影响,施工安全性更加难以保障。为解决以上问题,设计一种基于改进极限学习机的基坑底面下沉风险监测方法。引入松弛参数与抗弯刚度软化指数,得到载荷-平面移动分布弹性模量和裂缝开展度、钢筋屈服的内在关联,构建基坑底面下沉模型。构建建基坑底面下沉风险评估体系。利用极限学习机数据训练模型,计算延迟步数,划分风险监测临界值,完成基坑底面下沉风险监测训练。实验结果表明,对于降水阶段、注浆阶段以及开挖失水阶段,研究方法得到的地基下沉累计变化值与真实值具有更高拟合度,风险监测误判率可控制在5%以下。经过600次实验后,风险监测耗时保持在8ms~9ms,说明上述方法能够为建筑地基基坑挖掘施工安全保障提供可靠理论依据。
In order to guarantee the construction safety,this article designed a method for monitoring the subsid-ence risk of foundation pit based on improved limit learning machine.Firstly,By introducing relaxation parameters and flexural stiffness softening index,the intrinsic correlation between the elastic modulus of load plane movement distribution,crack propagation,and steel yield was obtained,and a foundation pit bottom settlement model was con-structed.Secondly,the risk assessment system of foundation pit bottom subsidence was established.Moreover,the limit learning machine data was used to train the model and thus to calculate the delay steps.Finally,the critical val-ue of risk monitoring was determined,thus completing the risk monitoring and training of subsidence.Following con-clusions can be drawn from experimental results.For the precipitation stage,grouting stage,and excavation water loss stage,the cumulative changes in foundation settlement obtained by the research method have a higher degree of fit with the actual value,and the risk monitoring misjudgment rate can be controlled below 5%.After 600 experiments,the risk monitoring time can be kept at 8ms~9ms,indicating that this method can provide a reliable theoretical basis for the construction safety of foundation pit excavation.
作者
余晓雅
张海清
YU Xiao-ya;ZHANG Hai-qing(College of Civil Engineering and Architecture,Hebei University,Hebei Baoding 071002,China)
出处
《计算机仿真》
北大核心
2023年第5期252-256,共5页
Computer Simulation
关键词
建筑地基
基坑下沉
风险监测
风险指标
改进极限学习机
Building foundation
Foundation pit sinking
Risk monitoring
Risk indicators
Improved limit learn-ing machine
