地铁结构变形预测的时间序列模型方法研究

Research of Time Series Model Method in Metro Structure Deformation Prediction

本文根据南京地铁某区间隧道结构变形监测数据,分别采用回归分析方法和时间序列分析方法建立模型,对地铁结构变形进行预测。计算结果表明在该工程实例中,回归分析模型的预测精度约为±0.20mm,时间序列模型的预测精度约为±0.08mm,较回归分析模型提高了60%,能够较好的对地铁结构变形进行预测。

The paper bases on deformation monitoring data of one tunnel＇s structure of Nanjing Metro,and establish model by using regression analysis method and time series method respectively,in order to predict the deformation of metro structure.In this engineering projects,the prediction precise of regression analysis model is about ±0.20 mm,while time series model＇s prediction precise is about ±0.08 mm,increasing 60% over regression analysis model.It shows that time series model can predict the deformation of metro structure better.