针对小样本、未知概率分布的变形数据粗差探测问题,提出了一种基于灰包络线的多维粗差探测方法,并根据灰色系统的不确定度理论,构造了粗差探测的灰包络线。将包络线探测法应用于某水电站的地下建筑物变形数据的粗差探测,验证了该方法是一种有效的粗差探测方法。
Aiming at the problem of the small samples whose distribution is unknown, a novel method using the grey envelope to detect the gross error in deformation monitoring data is proposed. Firstly, according to the uncer- tainty theory of grey system, the grey envelope for gross error detection is established, and then the principle and its steps for gross error detection are introduced. Using the method to detect the deformation in underground building of a hydropower station, we finally draw a conclusion that the proposed method is feasible and effective.