应用时间序列方法作大坝变形预报

Application of Time Series Analysis to the Prediction of Deformation for Large Dams

  • 摘要: 本文首先介绍时序分析的三个基本模型——ARMA模型、AR模型和MA模型,以及各模型的统计性质。然后以某大坝1715廊道的激光视准线观测位移值(已利用倒垂观测把相对位移化为绝对位移)为例,着重叙述大坝变形分析的建模过程,得到了一个AR(2)模型并对大坝变形作了预报,结果具有相当好的预报精度。从而说明,时序分析法将是大坝变形分析的一个有力工具。

     

    Abstract: This paper first introduces three basic models——the ARMA model, the AR model and the MA model in the time series analysis and their statistical properties. Taking the absolute displacements of a large dam in the western of China as an example, the paper puts the emphasis on the procedure of the establishment of the model for deformation analysis of the large dam, An AR(2) model has been built and used in the prediction of deformations.The result shows good prediction accuracy. Compared with the stepwise regression analysis, it is of better accuracy. Therefore, if the regression factors have not been observed, the method mentioned in this paper will show its advantages in deformation. analysis.

     

/

返回文章
返回