基于Vine-Copula的高土石坝变形监控模型研究
Research on deformation monitoring model for high earth-rock dams based on Vine-Copula
陈天赐 1李艳玲 1张芳 2陈枭3
作者信息
- 1. 四川大学 水力学与山区河流开发保护国家重点实验室,四川 成都 610065;四川大学 水利水电学院,四川成都 610065
- 2. 中国电建集团贵阳勘测设计研究院有限公司,四川成都 610065
- 3. 镇雄县水务局,云南 昭通 657200
- 折叠
摘要
针对高土石坝变形监控模型主要基于单一测点,无法定量考虑测点空间关联性的问题,构建了基于Vine-Copula的变形监控模型,并提出了基于蒙特卡洛随机抽样法的空间置信域预警阈值设置方法.在模型构建过程中,充分考虑了多测点之间的时序特性和空间相关性,利用Vine-Copula方法对变形数据进行精确建模和分析,以揭示高土石坝变形的整体趋势.同时,通过蒙特卡洛随机抽样法确定了空间置信域,为预警阈值的设定提供了科学依据.工程实践表明:该模型模拟结果能够准确反映高土石坝变形的整体趋势,具有较高的合理性和精度,有效实现了监测效应量向高土石坝空间全域的拓展.研究成果可为高土石坝安全监控提供新的思路和方法,具有重要的理论意义和实践价值.
Abstract
Aiming at the problem that the deformation monitoring model for high earth-rock dams is mainly based on a single measuring point and cannot quantitatively consider the spatial correlation of measuring points,a deformation monitoring model based on Vine-Copula was constructed,and a spatial confidence domain early warning threshold setting method based on Monte-Carlo random sampling method was proposed.In the process of the model construction,the time series characteristics and spatial correlation between multiple measuring points were fully considered,and the Vine-Copula method was used to accurately model and analyze the deformation data to reveal the overall trend of high earth-rock dam deformation.At the same time,the spatial confidence region was determined by Monte-Carlo random sampling method,which provided a scientific basis for the setting of early warning threshold.The engineering practice showed that the model simulation results can accurately reflect the overall de-formation trend of high earth-rock dam,with high accuracy and rationality,and effectively realize the expansion of monitoring effect-quantity to the whole space of high earth-rock dams.The research results can provide new ideas and methods for the safety monitoring of high earth-rock dams,and have important theoretical significances and practical values.
关键词
高土石坝/变形监控模型/Vine结构/Copula函数/空间置信域Key words
high earth-rock dam/deformation monitoring model/Vine structure/Copula function/spatial confidence domain引用本文复制引用
基金项目
国家重点研发计划(2018YFC0407103)
出版年
2024
国家科技期刊平台
NSTL
万方数据