Displacement monitoring model of concrete dams using the shape feature clustering‐based temperature principal component factor

Abstract

The mathematical monitoring model-based interpretation of recorded quantities, especially of displacements, is essential for the structural health diagnosis of concrete dams. In practice, dam displacements are frequently interpreted and predicted by the hydraulic, seasonal, and time model, which considers the thermal deformation effect of a dam body by the periodic harmonic factor. The main purpose of this paper is to replace this factor with the measured temperatures of a dam body. This approach is conducted by performing a time series shape feature-based spatial clustering method for the temperature field of a dam body in the first step. The principal components are then extracted from each cluster and used as the temperature factors in the monitoring model. An engineering example of the Jinping-I arch dam demonstrates the good performance of the proposed clustering method and established monitoring models. By comparing the shape feature clustering-based temperature principal component factor with the periodic harmonic factor, it can be concluded that the proposed models can describe the thermal deformation effect of concrete dams more reasonably: in the presented case study, especially for the dam parts of one half of the dam height above the foundation plane.