The unique structures and foundations of a dam make its safety monitoring complex task. As the most intuitive effect dams, deformation contains important information on evolution. Actual response has purpose diagnosis early warning compared with model prediction. Given poor generalization ability conventional statistical model, establishing is thus essential. prediction concrete using random forest regression (RFR) studied. To build an optimized RFR used to establish input variables, select...

Deformation monitoring is the main program in area of dam safety. Because statistical model simple and intuitive, it widely used safety monitoring. However, dam's displacement statistic model, there a high degree linear relationship between influence factors. Due to multicollinearity, models calculated with traditional methods are not accurate stable. Besides, because integrity, each part interrelated interactive. Currently, single point or multipoints cannot accurately reflect actual...

Grasping the change behavior of dam foundation seepage pressure is great significance for ensuring safety concrete dams. Because environmental complexity location, prototypical data are easy to be contaminated by noise, which brings challenges accurate prediction. Traditional denoising methods will lose detailed characteristics objects, resulting in prediction models with limited flexibility and accuracy. To address these problems, noise denoised using variational mode decomposition...

Dams are the main water retaining structures in hydraulic engineering field. Safe operations of dams important foundations to ensure functionalities these structures. Deformation, as most intuitive feature dams' operation behaviors, can comprehensively reflect dam structural states. In this case, analysis prototype deformation data and establishment a real-time prediction model become frontier research contents field safety monitoring. Considering multi-nonlinear relationships between...

Deformation is the most intuitive reflection of comprehensive behavior concrete dams; it great significance to predict and interpret deformation observation data for dam health monitoring. The world's highest dam, Jinping I arch in China, was discussed this paper. Aiming at its annually measured continuous growth phenomenon body towards downstream direction when reservoir keeps stable normal water level 1,880.0 m, influences cement hydration heat-induced temperature rise effect, valley...

In order to discover anomalies of dam structure behaviors and evaluate the operation status timely, it is quite demanding analyze safety monitoring data that has been collected from instruments. However, outliers in original may affect accuracy performance assessment, which need be detected before analyzing data. Model-based methods have applied outlier detection as a kind common method for long time, but they generally rely heavily on model easily lead misjudgment once complex. Considering...

Constructing an accurate dam displacement health monitoring (DHM) model is crucial to ensure the safety of dam. However, previous studies on DHM focused analysis and prediction a single measurement point, with little work multiple points, which leads low efficiency in evaluating overall status dams. Furthermore, majority these models are based hydraulic engineering moderate climatic areas, results accuracy when applied severely cold regions. To address issues, HT c T proposed full...

Real and effective monitoring data are crucial in assessing the structural safety of dams. Gross errors, resulting from manual mismeasurement, instrument failure, or other factors, can significantly impact evaluation process. It is imperative to eliminate such anomalous data. However, existing methods for detecting gross errors concrete dam deformation often focus on analyzing a single effect quantity. This lead sudden jumps values quantity caused by changes environmental variables being...

Seepage monitoring is a vital task in the risk management of concrete dams. Considering lag effect input factors, this paper presents novel seepage model for dams and proposes an effective identification method process. Firstly, extreme gradient boosting (XGBoost) were adopted to predict dam seepage. Hybridizing grey wolf optimization (HGWO) which integrates differential evolution (DE) into (GWO) five-fold cross validation utilized optimize hyper-parameters XGBoost. Secondly, under same...

Prediction models are essential in dam crack behavior identification. Prototype monitoring data arrive sequentially safety monitoring. Given such characteristic, sequential learning algorithms preferred over batch as they do not require retraining whenever new received. A methodology using the genetic optimized online extreme machine and bootstrap confidence intervals is proposed a practical tool for identifying concrete behavior. First, adopted to build an prediction model of The...

The safety of a high concrete arch dam should be rapidly diagnosed from different angles. Displacement is an actual comprehensive reflection the dam, and it very important to diagnose overall deformation behaviour by displacement-based mathematical monitoring models. In this article, based on spatial association validation measured displacement two dams empirical orthogonal function decomposition Pearson correlation analysis, association–considered models were proposed for multimonitoring...

The mathematical monitoring model-based interpretation of recorded quantities, especially displacements, is essential for the structural health diagnosis concrete dams. In practice, dam displacements are frequently interpreted and predicted by hydraulic, seasonal, time model, which considers thermal deformation effect a body periodic harmonic factor. main purpose this paper to replace factor with measured temperatures body. This approach conducted performing series shape feature-based...

As an important feature, deformation analysis is of great significance to ensure the safety and stability arch dam operation. In this paper, Jinping-I with a height 305 m, which highest in world, taken as research object. The data representation method analyzed, processing spatiotemporal discussed. A hybrid model established, hydraulic component calculated by finite element method, other components are still statistical method. Since relationship among measuring points not into account...

The unique structure of a dam complicates safety monitoring. Deformation can provide important information about evolution. In contrast to model prediction, actual response monitoring data be used for diagnosis and early warning. Given the poor mining ability conventional methods, it is essential develop method extracting factors influencing dam. this study, evaluation concrete deformation were developed using evidence theory random forest. has advantages being easily understood,...