基于BFAST时间序列分解与分段建模的混凝土坝多测点变形定量分析方法  

Quantitative Analysis Method of Multi-measuring Point Deformation of Concrete Dam Based on BFAST Time Series Decomposition and Subsection Modeling

作　　者：郭张军 陈容 GUO Zhang-jun;CHEN Rong(Power Research Institute of State Grid Shaanxi Electric Power Company Limited,Xi'an 710054,China;Datang Sichuan Power Generation Co.,Ltd.,Chengdu 610031,China)

机构地区：[1]国网陕西省电力有限公司电力科学研究院,陕西西安710054 [2]大唐四川发电有限公司,四川成都610031

出　　处：《水电能源科学》2024年第9期130-133,216,共5页Water Resources and Power

摘　　要：大坝运行环境复杂,其变形与外部荷载之间往往存在明显的非线性关系。为此,提出一种基于BFAST时间序列分解与分段建模的混凝土坝多测点变形定量分析方法,基于重力坝垂线监测数据,首先使用主成分分析提取多个测点的综合位移;其次使用BFAST时间序列分解方法将综合位移分解为季节性和趋势性两部分;然后通过合理选择改变点,建立分段统计回归模型,分时段量化各环境量对大坝位移的影响;最后通过对安康水电站水平位移规律的分时段定量分析验证了所提方法可行、有效。研究结果为更好地分析大坝变形性态演变历程和定量解释大坝变形机理提供了有力的技术支持。The operation environment of dams is complex,and there is an obvious nonlinear relationship between the deformation and the external load.This paper presents a quantitative analysis method of concrete dam deformation at multiple measuring points based on BFAST time series decomposition and sectional modeling.Based on the vertical monitoring data of gravity dams,principal component analysis is firstly used to extract the comprehensive displacements of multiple measuring points.Then,the BFAST time series decomposition method is used to decompose the comprehensive displacement into two parts:seasonal and trend.Finally,through the reasonable selection of change points,a segmented statistical regression model is established to quantify the impact of various environmental variables on dam displacement in different periods.The feasiblity and effectiveness of the proposed method is validated by the segmented quantitative analysis of the horizontal displacement law of Ankang Hydropower Station.The research results provide strong technical support for better analysis of the evolution of dam deformation behavior and quantitative interpretation of dam deformation mechanism.

关 键 词：BFAST 时间序列分解 突变检测 大坝安全监控 主成分分析 

分 类 号：TV698.1[水利工程—水利水电工程]