用统计模型估计大坝实际观测精度

观测精度历来是大坝观测中最需要关注的要素之一.在仪器埋设后,怎样检验设施的实际精度?目前采用现场率定的方法是必要和有效的,但存在一定的局限性.充分利用各工程已有的长期观测资料,根据其统计模型估计实际观测精度,是一种可普遍实行的有效方法.文中提出了衡量回归效果优劣的参考标准,并以垂线为例,对大量实测资料的模型进行筛选,分别提出了在人工观测和自动化监测条件下的量测精度和系统综合精度.在自动化监测资料分析中发现,环境的日变化对结构性能会有一定影响,目前统计模型中没有反映此类因素的因子,拟合效果会有所降低.

Estimating the Actual Accuracy of Dam Observation with a Statistical Model

Observation accuracy is one of the most important factors in dam observation. The key point is the checkout of the actual observation accuracy of the embedded instruments. Locale calibration of the instruments is necessary and effective, but has limitation. Estimating the actual observation accuracy with a statistical model of long-term observation data from constructed dams is an effective method, which can be widely used in future. The standard for judging the regression effect of the model is proposed. The measurement accuracy and integration accuracy of a plumb-line are given for manual observation and automatic monitoring, through filtering the models of a large quantity of observation data. It is found that the daily variation of the environment has some influence on the structural behavior of dams. None of these factors are applied to the present statistical models; as a result, the goodness of fit with be lowered.