摘要
为实现天津地区基坑变形影响因素筛选及预测,以天津6号线地铁基坑为工程实例背景,先利用数量化理论Ⅲ筛选各因素对基坑变形的影响程度,再利用多种单项优化预测模型构建出基坑变形组合预测模型,以实现基坑变形的高精度预测。实例研究表明:基坑变形的重要因素是基坑宽度、支护刚度和插入比,次要因素是基坑长度、基坑深度、典型软弱层厚度和一般软弱层厚度,而抗隆起稳定性属一般因素,且通过耦合度分析表明,抗隆起稳定性因素虽为一般因素,但其与其他因素间的耦合度较高,对基坑变形的影响不可忽略。同时,通过组合预测,不仅能进一步提高预测精度,还能提高预测结果的稳定性,所得组合预测结果的相对误差均小于2%,具有较高的预测精度,验证了该组合预测思路的有效性。
In order to realize the screening and prediction of the influencing factors of foundation pit deformation in Tianjin Area,the foundation pit of the No.6 subway in Tianjin is taken as the actual engineering case and the influencing degrees of all the factors are screened with the third theory of quantification at first,and then the combined prediction model for the foundation pit deformation is established with several single optimization prediction models,so as to realize the high-precision prediction of the foundation pit deformation.The case study shows that the important factors of the foundation pit deformation are foundation pit width,support stiffness and insertion ratio,and the secondary factors are foundation pit length,foundation pit depth,typical soft layer thickness and general soft layer thickness,while the heave-resistant stability belongs to a general factor.Moreover,it is indicated through the relevant analysis on the coupling degree that though the heave-resistant stability is a general factor,it has a higher coupling degree with other factors,thus its influence on foundation pit deformation cannot be neglected.Meanwhile,the prediction accuracy cannot only be further improved,but the stability of the prediction result can also be enhanced through the combined prediction,from which the relative error of the result from the combined prediction is less than 2%with higher prediction accuracy,and then the effectiveness of the idea of the combined prediction given herein is verified as well.
作者
王飞
WANG Fei(Shaanxi Railway Institute,Weinan 714000,Shaanxi,China)
出处
《水利水电技术（中英文）》
北大核心
2021年第2期127-133,共7页
Water Resources and Hydropower Engineering
关键词
地铁
基坑
影响因素
相关性准则
组合预测
subway
foundation pit
influencing factors
correlation criteria
combined prediction
