基于小波的时间序列分析在变形监测中的应用

Application of Time Series Analysis in Deformation Monitoring Based on Wavelet

变形监测的目的不再只是对变形现象进行观测,更重要的是对形变数据加以处理、分析并建立预测模型对后期的形变量做出合理的预报,减少建筑物因变形导致的事故。由于受到环境和设备自身因素的影响,变形监测数据采集过程中不可避免的会产生噪声数据,导致预测不准确。为了对后期的变形作出更准确的预测,本文采取了小波理论对实测数据进行降噪处理,再运用时间序列预测方法建模预测后期的变形数据并与后续的实测数据进行对比验证。

The purpose of deformation monitoring is not only to observe the deformation,but also to analyze and deal with the deformation data and establish a prediction model to make a reasonable prediction for the later shape variables so as to reduce the accidents caused by the deformation of the building.Due to the influence of environment and equipment itself,noise data will inevitably be generated during the deformation monitoring data acquisition process,resulting in inaccurate prediction.In order to make more accurate prediction for the later deformation,the wavelet theory is adopted to deal with the measured data noise,and then the time series prediction method is used to model the later deformation data and compare with the subsequent measured data.