基于TGM-ARMA模型的地铁隧道结构沉降预测分析

Prediction and analysis of structure settlement of the metro tunnel based on TGM-ARMA model

地铁隧道结构沉降主要受土层状况、地下水位、轨道荷载及周边环境等诸多因素影响,机理复杂,随机性强,而且随时间不断变化,难以用固定的数学模型表示,是一个非平稳的过程。据此,结合以灰色系统理论为基础的TGM(1,1)模型和以时间序列为基础的ARMA(n,m)模型各自的特点,建立了时变参数灰序模型TGM(1,1)-ARMA(n,m),并对地铁隧道主体结构沉降进行了分析与预报,结果表明时变参数灰序模型预测精度较高,且适用于中长期预测。因此,该组合模型在地铁隧道结构沉降预测分析中有较强的适用性,具有一定的应用价值。

The structure settlement of metro tunnel is mainly affected by status of the soil,groundwater level,track load,surrounding environment and many other factors.Their complicated mechanism,strong randomness,time-related variation all make it hard to describe by one fixed mathematical model,therefore,the structure settlement of metro tunnel is a non-stationary process.Accordingly,combining the characteristics of TGM（1,1） model based on grey system theory and ARMA（n,m） model based on time series analysis,a time-varying parameter gray model of TGM（1,1）-ARMA（n,m） was established.The application of the model to analysis and predict of the main structure settlement of metro tunnel shows that the prediction precision is high and the model is suitable for middle long-term prediction.Therefore,the combined model has relatively strong applicability in analyzing the structure settlement of metro tunnel.