摘要
针对传统大坝变形异常值识别方法准确性低、可靠性差及难以程序化实现等问题,以锦屏一级大坝为例,基于趋势面分析方法构建了大坝变形非线性统计模型,以生成的三维趋势面为基准,通过设置容许误差构造了异常值识别包络面,并在计算机系统层面设计了程序实现路径,针对异常值的产生原因及数据特征,提出了融合系统远程召测、采集设备状态自检及设定初、复测值欧氏距离判别条件的异常值性质判别方案。结果表明:该方法无需动态计算即可实现高准确度的变形异常值动态识别,所生成的三维趋势面具有明确物理力学意义,能精准评估不同库水位与环境温度条件下的大坝变形性态。
To address the issues of low accuracy,poor reliability,and difficulties in programmatic implementation associated with traditional methods for identifying outliers in dam deformation,using the JinpingⅠDam as a case study,a nonlinear statistical model for dam deformation was developed based on trend surface analysis.Using the generated three-dimensional trend surface as a benchmark,an outlier identification envelope was constructed by setting allowable errors,and a program implementation pathway was designed at the computer system level.In response to the causes and data characteristics of outliers,a comprehensive solution for determining the nature of outliers was proposed,integrating remote system recall testing,self-diagnosis of data acquisition equipment status,and Euclidean distance criteria between initial and repeated measurements.The results demonstrate that the proposed method achieves highly accurate dynamic identification of deformation outliers without requiring dynamic computation.The generated three-dimensional trend surface possesses clear physical and mechanical significance,enabling precise evaluation of dam deformation behavior under varying reservoir water levels and ambient temperatures.
作者
刘健
王继敏
杨强
张晨
崔岗
LIU Jian;WANG Jimin;YANG Qiang;ZHANG Chen;CUI Gang(Yalong River Hydropower Development Company Limited,Chengdu 610051,China;School of Civil Engineering,Tsinghua University,Beijing 100084,China;State Grid Electric Power Research Institute/NARI Group Corporation,Nanjing 211106,China)
出处
《水利水电科技进展》
北大核心
2025年第6期77-84,共8页
Advances in Science and Technology of Water Resources
基金
国家自然科学基金重点项目(51739006)。
关键词
大坝安全
在线监控
变形监测
异常值识别
锦屏一级大坝
dam safety
online monitoring
deformation monitoring
outlier identification
JinpingⅠDam
