The arch dam deformation monitoring model is the most commonly used method for arch dam health monitoring. Aiming at the deformation monitoring problem of extra-high arch dams, this paper proposes an intelligent optimization support vector machine deformation monitoring model for concrete extra-high arch dams. Particle swarm optimization (PSO) was used to optimize the penalty factor, kernel function parameters of the support vector machine (SVM), and tolerate bias. The deformation monitoring model of concrete extra-high arch dam based on PSO-SVM was established, and the influence of aging factors on the model performance was analyzed. Engineering examples show that the PSO-SVM deformation monitoring model of concrete extra-high arch dam has good prediction accuracy and generalization ability, which is suitable for deformation monitoring of extra-high arch dam.

In order to reduce the impact of bridge pier groups on the water flow at the intersection of the main rivers of Hangzhou Zhijiang New City, three shoreline optimization schemes were studied to improve the flow pattern based on a two-dimensional mathematical model. The optimal remediation plan was recommended based on multiple objectives such as water resistance ratio, land area, water area, bridge pier safety, and so on. The results show that the widening one side of the river channel to prevent bridge piers from being located in the main channel or changing the route to avoid bridge piers can effectively reduce the impact of bridge piers on the water level. The three optimization schemes have the same improvement effect on the backwater of the bridge piers. Scheme 2 and scheme 3 will affect the safety of bridge piers, with relatively small water area and large land acquisition area. Scheme 1 has little impact on the safety of bridge piers and is relatively easy to implement. Scheme 1 has high comprehensive advantages, and is recommended as the optimal option.

In view of the local large deformation, collapse and erosion of silty mudstone entrance section of Daliangshan Highway #2 tunnel, through the field monitoring, indoor swelling test of surrounding rock and 3D modeling analysis, the mechanism of large deformation of initial support was studied. The impact of factors including rainwater, secondary deformation of surrounding rock and silt bias on the large deformation was discussed. The remediation plan was put forward. The results show that the surface clay rock above the tunnel is swelling rock, and the silty mudstone is non-expansive rock, and the tunnel deformation has little affected by the expansion of the surrounding rock. Affected by continuous rainfall, the change in the settlement of the vault is the most obvious, and the settlement increased by 150 mm compared with the pre-rainfall. The theoretical value of horizontal pressure on the deep buried side of the tunnel is 1.73 times that of the shallow side, and the tunnel structure is subjected to obvious bias load. The comprehensive treatment scheme of strengthening temporary support + advanced support of large pipe shed + grouting reinforcement of double-layer small conduit + addition of anti-slip piles at the junction of light and dark at the top of the cave + two rows of pipe piles on the left side of the left hole and two rows of pipe piles in the middle of the left and right holes of the hole section of the hole was adopted, and the deformation of the tunnel was reduced by 87%, and the large deformation of the initial support is effectively controlled.

Anti-seepage wall is an important anti-seepage structure in cofferdam, which plays an active role in improving the distribution of seepage field and the stability of cofferdam. This paper presents an optimization method of cutoff wall structure based on artificial fish optimization algorithm. Taking a cofferdam project as an example, the calculation model of cofferdam was established based on finite element analysis, and the wall thickness and rock depth of cutoff wall were taken as the variables to be optimized. Finally, the safety factor of cofferdam under the condition of constant water level and sudden drop of water level was obtained through seepage and stress coupling analysis. Compared with the traditional artificial fish swarm algorithm, the results show that the optimized method based on the improved artificial fish swarm algorithm has faster convergence speed and convergence accuracy, and the optimized anti-seepage wall structure can achieve the balance between the anti-seepage effect and the economy. The pore water pressure behind the cutoff wall decreased significantly, and the safety factors of the cofferdam under the conditions of constant water level and falling water level (falling rates are v=0.5 m/d, 1.0 m/d and 2.0 m/d, respectively) were 1.489, 1.410, 1.376 and 1.321, which met the safety requirements, reflecting the feasibility and rationality of the proposed method. The research results can provide reference for the design and construction of cofferdam.

Using GRACE gravity satellite data combined with GLDAS data inversion can obtain the change of groundwater reserves, which can monitor the change of groundwater reserves in the study area on a large scale, but the spatial resolution of the obtained data is only 0.25°, and it is difficult to be applied in the small scale. Based on the spatial relationship between GRACE terrestrial water storage and GLDAS shallow surface water storage and precipitation, NDVI, a downscaling method of groundwater storage change based on geographically weighted regression model (GWR model) was proposed, and the spatial resolution of groundwater storage was downscaled to 1 km. The results show that the GWR model downscaling method is successfully applied to the Haihe River Basin. The correlation coefficients between the groundwater storage data after downscaling and the measured groundwater level data at 46 verification points are all greater than 0.6, and the simulation results are reasonable and reliable. Furthermore, the temporal and spatial variation characteristics of groundwater storage in Haihe River Basin and Beijing Plain were analyzed by using the downscaling results. The results is highly consistent with the existing data, indicating that the downscaling method based on GWR model can effectively improve the spatial resolution of groundwater storage change data.

Analyzing the spatial and temporal characteristics of extreme precipitation and the response of wind speed before and after extreme precipitation are of great significance for extreme precipitation characterization, improving the understanding of the relationship between extreme precipitation and wind speed, as well as managing water-related disasters. This study defined event-based extreme precipitation by considering preceding and succeeding precipitation, and classified precipitation patterns into three types, i.e. early peak, later peak and bimodal types according to the locations of precipitation peak, after which the spatiotemporal distributions of extreme precipitation and its relationship with wind speed were analyzed. The results show that single-peak of extreme precipitation occurs more often than bimodal one, and the amount increases from northwest to southeast, with significant upward trend of extreme precipitation detected; Extreme daily precipitation threshold increases with higher wind speed over the southeast and northeast, contrary to the southwest; Moreover, wind speed tends to show anomalies before and after extreme daily precipitation, and it decreases when extreme daily precipitation occurs over the southwest; However, over other regions, it increases before extreme daily precipitation and decreases after then, reaching the peak at the extreme daily precipitation day.

In order to quantitatively analyze the impact of upstream reservoirs and the operation of the Three Gorges Project on the hydrological rhythms of the lower reaches of the Yangtze River and lakes, M-K test, cumulative anomaly method, wavelet coherence spectrum method and other hydrological time series trend analysis methods with STL time series decomposition method were used to analyze the 49-year historical water level data of Sheshan Station at Lake Shijiu. The long-term change trend and law of hydrological rhythm of Lake Shijiu under the new relationship between Yangtze River and Lake Shuijiu were analyzed. The coherent relationship between hydrological rhythm changes and local meteorological factors were analyzed along with the ecological effects under the new river-lake relationship. Under the new river-lake relationship, the dry season is earlier and longer and the averaged water level is higher, the extreme low water level is more likely to occur, the variation is gentler, and it is difficult for high water level to happen. The extreme high water level in the flood season is lower and happens later, the averaged water level is lower, ends earlier, and the amplitude is gentler. The early dry season is beneficial to increase aquatic plant biomass, but the decrease of water level variation in dry season may cause biodiversity to decrease.

Currently, the researchers have not yet conducted a detailed analysis of the anchor block stability and there exist shortcomings in the discriminatory formula for the slip length. On the basis of the existing computational formulas, the axial force formula for the pipeline was optimized considering expansion joints, and the axial force formula was proposed for downstream elbow pipe without considering the expansion joints. Furthermore, by analyzing axial thrust on pipelines in the context of considering or not considering the expansion joints and comparing the concrete amount of the anchor block under all conditions, the anchor block of the buried steel pipe was found to have the minimum amount of concrete used in the context of not considering the expansion joints and under certain condition, which can provide a reference for future research and design work.

The noise and nonlinear characteristics in the deformation sequence of concrete dam seriously affect the accuracy of dam deformation prediction. In this paper, ensemble empirical modal decomposition (EEMD) was used to decompose the horizontal displacement signal of the dam to mine the effective deformation information. The singular spectrum analysis (SSA) was used to extract features from the high-frequency eigenmodal components (IMF) obtained from the decomposition to reduce the loss of effective information. Considering the complex stochastic and non-linear mapping relationship between effector and environmental variables, extreme gradient boosting (XGBoost) was used to model the prediction of the noise-reduced data. Considering the significant influence of XGBoost hyperparameters on the prediction performance of the model, the Northern Goshawk algorithm (NGO) with better global search capability was introduced to perform parameter search, and an NGO-XGBoost-based dam displacement prediction model was constructed. The calculation results show that the EEMD-SSA can effectively remove the noise from the dam displacement monitoring information, and the dam deformation prediction model based on NGO-XGBoost can significantly improve the prediction accuracy.

In order to study the impacts of aggregate size and fiber length on the mechanical properties of basalt fiber concrete, three aggregate sizes and basalt fiber lengths have been taken into account, and 108 concrete test blocks (including 72 cubic test blocks and 36 prism test blocks) have been designed and made. Furthermore, axial compression test, splitting tensile test as well as bending test have been carried out to analyze each parameter influence on compression strength, splitting tensile strength and bending strength. On the basis of random fiber script, ABAQUS finite element modeling analysis is carried out, and the corresponding strength calculation has been proposed. The results show that with the increase of aggregate size, the mechanical properties of matrix concrete are improved. While with the adding of basalt fibers, the concrete mechanical properties rise higher than aggregate size increase. The concrete mechanical properties present good performance when 30mm aggregate size mixes with 18 mm and 24 mm basalt fibers. The result of designed random fiber script in ABAQUS finite element modeling matches with the experiment well, and the calculation method considering two-parameter influence of aggregate size and basalt fiber length has good applicability.