时间序列在变形监测中的应用

Application for Deformation Monitoring of Time-series Analysis

详细介绍了时间序列的基本原理以及时间序列建模中的数据预处理、模型选择、模型定阶、模型参数的确定及模型的可靠性等建模步骤,并给出了时间序列的预报公式。最后用时间序列分析处理了某市的某一边坡沉降观测数据。通过对结果的分析,得出时间序列分析用于简单的时序预测具有较好的效果,9期预测误差的绝对值均小于0.18 mm。其建模过程与其它的数据处理方法相比,较为简单,是处理动态数据的一种非常有效的方法。

In this paper,the basic principle of time series analysis and the procession of it using in the deformation were introduced first.The modeling procedure of time series analysis was detailed intrduced,which contions data pre-processing,Selection of the Model,order estimation,confirmation of model parameters and the reliability.Two formulars for forecasting of ARMA and ARIMA are briefly introduced in this paper.At last,A slope observation of a city was tackled using the time series analysis.The result shows that the precision of the time series analysis untilyling in the stackling of some simple deformation series is very high.In addition the error absolute values of the nine periods were under 0.18 mm.At the same time,compared with other methods of data handling,the modeling procedure of time series analysis is simpler.It is an effective way to handle the dynamic data.