大坝位移监测中的粗差剔除方法

发布时间：2018-04-17 14:31

本文选题：大坝位移监测 + 粗差剔除　；参考：《水力发电》2017年09期

【摘要】：针对大坝位移监测中常用粗差剔除法存在的问题提出了一种新型的粗差剔除方法。首先,根据层次分析法得到Dixon准则、Grubbs准则、Chauvenet准则、四分点准则的权重系数,并设定好粗差判定阈值;其次,根据原始观测资料建立大坝位移的偏最小二乘回归模型,得出样本对应的残差序列;然后根据四个准则依次判定出所有疑似粗差的测值;最后,按照此4种准则的权重系数,加权综合得到每个测值的粗差概率值,并与粗差判定阈值对比,得到最后的粗差集并剔除。该方法综合考虑了多种粗差数据识别方法,权衡了误判和漏判的可能性,为剔除大坝位移数据中的粗差提供了一种更为科学合理的方法。
[Abstract]:A new method for eliminating gross errors in dam displacement monitoring is presented in this paper.Firstly, according to the analytic hierarchy process (AHP), the weight coefficients of the Dixon criterion and the quadrature criterion are obtained, and the threshold of gross error is set up. Secondly, the partial least square regression model of dam displacement is established based on the original observation data.Finally, according to the weight coefficients of the four criteria, the gross error probability value of each measured value can be obtained by weighted synthesis.Compared with the threshold of gross error, the final rough error set is obtained and eliminated.This method considers a variety of gross error data identification methods, and balances the possibility of misjudgment and omission, and provides a more scientific and reasonable method for eliminating the gross error in dam displacement data.
【作者单位】：南瑞集团公司(国网电力科学研究院);
【分类号】：TV698.11

【相似文献】