摘要
以某隧道工程基坑为例,给出了30日基坑支护结构26个监测点的竖向位移与水平位移以及64个道路表面点竖向位移的监测数据。在支持向量机(SVM)算法基础上,利用人工鱼群算法进行全局参数寻优,建立数据预报的优化SVM算法。通过实测监测数据的预报分析与比较,评估了优化算法的预报精度。结果表明:优化算法对3种类型监测数据的预报精度改进率分别达36.7%,37.8%和42.2%。
Taking a hydropower project foundation pit as an example,this paper discusses the main contents of foundation pit safety monitoring,and gives 30 days’monitoring data of vertical and horizontal displacement of 26 monitoring points of foundation pit supporting structure,and vertical displacement of 64 road surface points.Based on the support vector machine(SVM)algorithm,the artificial fish swarm algorithm is used for global parameter optimization,and the optimized SVM algorithm for data prediction is established.The results show that the accuracy improvement rate of the three types of monitoring data is 36.7%,37.8%,and 42.2%,respectively.
作者
王鸿
梁加俊
Wang Hong;Liang Jia-jun
出处
《建筑技术开发》
2024年第2期158-161,共4页
Building Technology Development
关键词
基坑
变形监测
SVM算法
位移
foundation pit
deformation monitoring
SVM algorithm
displacement
