时间序列分析在隧道施工监测中的应用

APPILICATION OF TIME SERIES ANALYSIS METHOD TO MEASURED DISPLACEMENT IN TUNNELING

用时间序列分析的更一般的组合模型描述隧道施工监测结果的变化规律，可以提高拟合模型的精度和预报的可靠性。其中，数据序列随时间变化的均值可用多项式或指数函数描述，而零均值的平稳过程可用ＡＲ（ｎ）或ＡＲＭＡ（ｎ，ｍ）模型拟合。对设测点前的断面位移值利用组合模型进行后推，并通过工程实例对该模型作了初步检验。

A more general time series associate model is used to describe the variation regularity of monitoring results for tunneling. The accuracy and reliability of displacement prediction can be improved with this model. In this model, the mean of data variation with time can be described by polynomial or exponential expression, and the stable process of zero mean by random process model AR(n) or ARMA(n, m) . The release displacement of tunnel before the first measurement is estimated with this model, and the method is checked by engineering example.