Segmented modeling method of dam displacement based on BEAST time series decomposition

Abstract

The dam and its operating environment are complex and uncertain and often show many different operating rules during its operating period. In this paper, firstly, the principal component analysis is used to extract the comprehensive displacement of multiple measuring points; Then, the comprehensive displacement is decomposed into seasonal and trend parts by using the Bayesian Estimator of Abrupt change, Seasonal change, and Trend (BEAST) method, and the change of its change law is quantitatively analyzed; Finally, the influence of each environmental quantity on the dam displacement is quantified by selecting the change point reasonably. The method proposed in this paper is verified by the quantitative analysis of the displacement law of multiple measuring points of an earth rock dam in different periods. The research results provide strong technical support for better analysis of the evolution process of dam deformation behavior and quantitative interpretation of dam deformation mechanism.