Pumped storage power stations usually arrange galleries in the backfill area at the bottom of the reservoir basin. Under the influence of uneven deformation, the galleries may be difficult to adapt to deformation and generate cracking, which can affect dam safety. In this study, the upper reservoir of Hohhot pumped storage power station was taken as a case study. Through a combination of monitoring data and numerical simulation, the deformation characteristics of the galleries on the backfill foundation were analyzed, and the causes and mechanisms of galleries cracking and structural joints damage were revealed. The in situ monitoring records cover the internal settlement of the dam, the deformation and seepage flow of the galleries, and the ambient temperature. Based on actual engineering data, a numerical model considering the structure and filling method of dam, backfill area, and gallery was established, and the calculation parameters of rockfill material constitutive model were inverted by the direct back analysis method. The monitoring data analysis and numerical calculations showed that the long length of the gallery and the sudden drop of the ambient temperature were the main reasons for the longitudinal microcracks in the top arch of the galleries in the backfill area; the strong constraint of bedrock and the uneven settlement of backfill foundation were the root causes for the penetrating cracks in the galleries at the junction of backfill area and bedrock. In addition, the depth of the gallery embedded in the bedrock determines the deformation form (torsional deformation or bending deformation) of the galleries at the junction of the backfill area and bedrock. Based on the monitoring and numerical simulation, the long-term deformation of the galleries and the development of structural joints were also predicted.

Constructing an interpretable model for the long-term deformation Structural Health Monitoring (SHM) of earthrock dams is of great significance for improving the safety state evaluation and monitoring effect. In this paper, a physics -data -driven model for the deformation SHM of earth -rock dams is proposed based on deep mechanism knowledge distillation. Firstly, the deterministic model is established based on the Finite Element Model (FEM) and outputs the hydraulic load component curve and aging component curve. Then a regression prediction model (HTSGAN) between influencing factors and deformation measurements at multiple measurement points is established based on the Graph Convolutional Network (GCN) and attention mechanism. Finally, the TeacherHydraulic -Time -Seepage Graph Attention Networks (T-HTSGAN) model is established based on the featurebased multi -teacher knowledge distillation using the knowledge of hydraulic loading physics and soil -rock creep physics of the FEM for mechanism constraints. The model effectively solves the problems of poor model interpretability and lack of physics knowledge constraints in previous earth -rock dam SHM models. The research results are applied to a project of a 185.5 -meter -high concrete -faced rockfill dam, and the predictive performance of the model is more effective and stable through the comparison of six baseline models. The comparative analysis of the component curves proves the effectiveness of the proposed knowledge distillation method for mechanism constraints and improves the interpretability of the neural network model. Therefore, the model is more suitable for engineering applications.