土岩组合地区基坑变形的预测方法研究

Study on Prediction for Excavation Deformation in Rock-soil Combination

对影响土岩组合地区基坑变形的主要因素进行了分析,并采用人工智能——BP神经网络的方法对土岩组合地区基坑的变形进行了预测研究。基于已有的研究资料,分析和总结了影响基坑变形的主要因素;建立了BP神经网络预测模型;借助MATLAB语言进行编程,利用训练稳定的网络模型,预测了基坑土岩组合地区基坑的最大侧移量。经与实测值比较,预测精度可满足工程的需要。该预测研究可对土岩组合地区基坑的设计与施工提供一定的参考。

Here, the main factors are detailedly analyzed, which would have a major effect on the excavation deformation in rock-soil combination. At the same time, the future deformation regularity of the excavation is predicted and researched by using the method of artificial intelligence-BP neural network. Based on the predecessors＇ research data, the main factors are analyzed and summarized. Then, the BP neural network model is built by using the samples＇ database. Finally, using the stable training model, the future deformation regularity is deduced by programing MATLAB procedure. Compared with the measured values, it is shown that the prediction values here could satisfy the civil engineering needs. On some extent, this study could provide some references for excavation design and construction.