时间序列分析方法在变形数据的处理中具有良好的模型拟合和预报精度。本文阐述了时间序列分析模型的时域特性,归纳了时间序列建模步骤,并结合监测工程的数据进行建模分析与预测。实例表明,时间序列分析方法所建立的模型拟合精度高,10期预测误差绝对值在0.3mm以内,能精确地反映建筑物的变形情况,从而避免或减少灾害的发生。 The time series analysis method has good model fitting and forecast accuracy in the processing of deformation data. In this paper, the time-domain characteristics of time series analysis model are expounded, the time series modeling steps are summarized, and the model analysis and prediction are combined with the data of monitoring project. The examples show that the time series analysis method has high fitting accuracy and less than 0.3mm absolute value of 10-stage prediction error, which can accurately reflect the deformation of buildings and avoid or reduce the occurrence of disasters.