线性回归在滑坡监测数据粗差判别中的应用

Application of linear regression theory in landslides monitoring gross error analysis

滑坡监测数据预处理中,须剔出含有粗差的数据,本文以马家沟滑坡相邻形变监测点多期监测数据相关性分析,从而建立一元线性回归方程,采用估算值中误差判定实测变形值是否存在粗差,并为滑坡稳定性分析提供了可靠的变形依据。

In pre-processing landslide monitoring data, the poor data should be rejected. This paper analyzed the correlation among multi-phase Majiagou landslide deformation monitoring data, and established a linear regression equation. Errors of the estimates were used to determine if there were gross errors in the measurements. So the monitoring data can providea reliable basia for landslide stability analysis.