大坝自动化监测数据的漂移校正

EXCURSION CORRECTION METHOD FOR AUTOMATIC MONITORING DATA OF DAMS

应用统计学原理,研究了对大坝自动化监测数据的区间性漂移进行校正的方法。首先,分别对测值序列中各段无漂移的子系列,以监测效应量(子系列效应量的平均值)的波动值作为因变量、以环境因子(子系列因子的平均值)的波动值作为自变量,逐段准备数据;之后,将各区段中准备好的数据融合在一起,进行全序列的回归分析;最后,根据回归分析的结果及漂移前后环境量的区段平均变化程度估计漂移量,并对测值进行校正,给出了具体的校正公式。此外,通过方差分析对校正精度进行了估计。

Based on the principle of statistics, the excursion correction method for automatic monitoring data of dams is researched. Firstly, for each sub-series without any excursion in the total motoring data series, data are prepared with the fluctuation of monitoring data as the dependent variable and the fluctuation of environment data as the independent variable. Then, the prepared data are fused for regression analysis. Finally, according to the regression results and the average fluctuation of the environment data of sub-series before and after excursion, the excursion can be estimated and corrected. Some correction formulas are put forward; besides, the precision of correction is estimated by variance analysis.