Abstract:Based on the remote sensing data of underlying surfaces and high-resolution DEM data in Shandong Province in 2020, the underlying surface factors affecting runoff were investigated, and the land surface type for runoff concentration was classified according to geomorphology-land use combinations. The land surface types for runoff concentration of 53 sub-basins with different areas in Shandong Province were analyzed. According to China’s 1∶1 000 000 digital geomorphological classification programme and considering the actual geomorphological situation in Shandong Province, the geomorphological types of Shandong Province were mainly divided into three types: plains, hills and mountains, with critical relief amplitude of land surface values of 30 and 100 m between plains, hills and mountains, and corresponding critical slope values of 1.8° and 5°. Based on the land use classification system of China’s multi-period land use remote sensing monitoring dataset, land use in Shandong Province was finally classified into eight major categories,with arable land being categorized into paddy field and dry filed, and bare land being listed separately. By combining the geomorphological and land use classification results, 22 land surface types for runoff concentration in Shandong Province were obtained, such as plain-dry field, plain-construction land, and hill-dry field, among which the area of plain-dry field accounted for the largest proportion of area. The average classification results show that as the sub-basin area increases, the number of land surface types for runoff concentration increases until it reaches stability, while the percentage of land surface type for runoff concentration with maximum area decreases until it reaches stability.

Abstract:Using the MIKE model and the ArcGIS platform, spatial information grids for the natural and social attributes of flood risks were respectively constructed and superimposed. The subjective and objective weights were calculated using the analytic hierarchy process and the entropy weight method, and the comprehensive flood risk index was computed to conduct a risk assessment for three adjacent streets in Qinhuai District of Nanjing City. The results show that, under different rainfall intensities, the spatial distribution of the comprehensive flood risk index is mainly characterized by higher values in the northwest and lower values in the southeast, with relatively higher values in vulnerable and sensitive areas. As the rainfall intensity increases, the area with higher values of the comprehensive flood risk index continuously expands. The main waterlogging treatment points in the study area are all located in regions with higher comprehensive flood risk indices. The flood risk assessment based on spatial information grids can provide references for more refined and dynamic flood disaster risk assessment and flood management.

Abstract:In order to more accurately determine whether cavitation occurs in the model hydroturbine, an intelligent identification method for turbine polymorphic images is proposed based on image recognition. This method extracts features from the target turbine runner images through machine preprocessing and binarization, and constructs the target feature matrix of the target image. The target feature matrix is multiplied by the correction value obtained from the expert database experience and is input into the cavitation identification model. It is then compared with the template correction feature matrix of the template images stored in the model to achieve cavitation identification of the water turbine runner. The practical application results of the project show that the identification accuracy of this method is about 80%, with a slight occurrence of false positives, but it can meet the requirements for practical use. Compared with existing methods, this method can not only improve the speed of turbine cavitation identification, but also enhance the identification quality, thereby achieving intelligent, mathematical, and simplified turbine cavitation detection.

Abstract:In order to analyze the deformation and failure mechanism of canal slopes, the consolidated drained and consolidated undrained triaxial shear tests were carried out on the collapsible loess of an open-channel water transfer project in Xinjiang. The triaxial test shear curve was simulated by the improved Duncan-Chang model and compared with those obtained from the Duncan-Chang model and the hump-type cubic curve model. The results show that with the increase of confining pressure, the stress-strain curve and volume change curve of consolidated drained test change from softening type to hardening type. The stress-strain curves of consolidated undrained test are all hardening type, and the pore pressure curves are all softening type. The effective cohesion obtained by the consolidated drained test is greater than that obtained by the consolidated undrained test, and the effective internal friction angle obtained by the consolidated drained test is slightly different from that obtained by the consolidated undrained test. There are nine parameters in the improved Duncan-Chang model. The Duncan-Chang model cannot reasonably describe the softening triaxial test curve, and the hump-type cubic curve model has a poor fit to the strong softening triaxial test curve, while the improved Duncan-Chang model can better describe various types of triaxial test curves of collapsible loess and has good accuracy.

Abstract:To gain an in-depth understanding of the relationship between capillary suction and moisture content in cement-based materials, a prediction model for capillary suction in cement-based materials considering pore fractal characteristics was developed based on the fractal geometry approach and the capillary theory. This model establishes a fitting relationship between the fractal model parameters and the sample mix proportion, enabling rapid prediction of capillary suction for cement-based materials with different mix proportions. Verification experiments measuring pore size distribution were conducted based on the relative humidity method. The results indicate that samples with different mix proportions (water, ash, and sand) exhibit significant fractal characteristics, which could be well described by the Menger sponge fractal model. The deviation between the predicted results of the capillary suction model and the experimental test results was mainly within ±20%, demonstrating that the model can accurately predict the relationship between capillary suction and moisture content in cement-based materials.

Abstract:To study the evolution law of the mechanical properties of cement mortar under the influence of the relative ice content in pores, based on the resistivity test results of cement mortar under periodic temperature variations, the time-history variation law of the relative ice content of cement mortar was calculated and analyzed. Using the PPR modeling technology, a calculation model for the relative compressive strength under the influence of different relative ice contents was established. The results show that under periodic temperature variations, the resistivity of cement mortar has a strong correlation with the relative ice content, and the established calculation model can effectively predict the changes in relative compressive strength with changes in the water-to-binder ratio, saturation, number of cycles, and maximum relative ice content of cement mortar under periodic temperature variations.

Abstract:To scientifically evaluate the effectiveness of waterlogging control in plain water network areas, taking Dianshanhu Town in Kunshan City as a study area, the MIKE FLOOD model, which integrates a one-dimensional river network, a two-dimensional terrain, and a one-dimensional drainage network, was used to simulate urban flooding in the plain water network area. The model was calibrated and validated using observed rainfall data, allowing for accurate simulations of maximum inundation depths in current situation and after treatment under four different rainfall return periods: 2-year, 5-year, 10-year, and 20-year events. The simulated results indicate that the inundation area changes significantly under the 20-year return period, while it shows no substantial variation under the other three return periods. For urban waterlogging control under smaller rainfall return periods, priority should be given to upgrading drainage networks, while for intense short-duration storms, the construction of sponge city facilities should take precedence. To address urban flood risks caused by various return periods and short-duration rainfall, a combination of drainage system upgrades and sponge city facilities is recommended.

Abstract:To improve the accuracy of daily runoff time series prediction and improve the prediction performance of the regularized extreme learning machine (RELM), the optimization performance of the improved dung beetle optimization (IDBO) algorithm and improved dwarf mongoose optimization (IDMO) algorithm was compared and verified, and the WPT-IDBO-RELM and WPT-IDMO-RELM models for daily runoff time series prediction were proposed based on wavelet packet transform (WPT). The daily inflows of the Mudihe Reservoir and Malutang Power Station in Yunnan Province were predicted. The results show that the average absolute percentage errors of the WPT-IDBO-RELM and WPT-IDMO-RELM models in predicting daily runoff for the Mudihe Reservoir are 1.048% and 1.015%, respectively, and 1.493% and 1.478% for Malutang Power Station, which are better than other comparative models. The optimization performance of the IDBO and IDMO algorithms on standard test functions and instance objective functions is better than that of comparative algorithms. The better the optimization performance of the IDBO and IDMO algorithms, the better the hyperparameters of the RELM, and the higher the prediction accuracy of the WPT-IDBO-RELM and WPT-IDMO-RELM models. WPT can decompose the daily runoff series into subseries components that have stronger regularity and are fewer in number, significantly reducing model complexity and computational scale while improving prediction accuracy.

Abstract:To address the challenge of obtaining high-resolution topographic data for urban flood simulation, a method for constructing high-precision terrain was proposed by integrating multi-source data, including drainage network data, road data, and building data. Taking the area surrounding the Longwangmiao River in Tangshan City as a case study validation area, an urban stormwater model (GAST-SWMM coupled model) was developed based on the digital elevation model (DEM) constructed by the proposed method, and was used to simulate the stormwater processes under two different rainfall events. The simulated waterlogging positions closely match the actual ones. For one of the rainfall events, the Nash-Sutcliffe efficiencies for the water levels at two manholes reach 0.92 and 0.87, respectively, with relative peak water level errors of 3.35% and 3.63%. These verify the effectiveness of the proposed method, indicating that the DEM constructed using this method can accurately reflect the topographic features of the study area and meet the accuracy requirements for urban flood simulations.

Abstract:Based on the long-term series data of water level, flow, and precipitation, a long short-term memory (LSTM) neural network model using the particle swarm optimization (PSO) algorithm (PSO-LSTM model) for hyperparameter optimization was constructed to provide short-term forecasts with forecast periods of 1 to 3 days for the water level during the flood season at Pingwang Station, the center of the plain river network in the Wujiang section of the Jiangnan Canal in Suzhou. The water level prediction results were compared with those from water level prediction models based on the PSO algorithm, including support vector machine (SVM), random forests (RF), convolutional neural networks (CNN), and gated recurrent units (GRU). The effects of water conservancy projects on the prediction accuracy of water level were investigated. The results show that the PSO-LSTM model has high prediction accuracy for short-term forecasts with forecast periods of 1 to 3 days, but the prediction accuracy gradually decreases with the growth of the forecast period. Compared with the PSO-SVM, PSO-RF, PSO-CNN, and PSO-GRU models, the PSO-LSTM model has a lower mean absolute percentage error and better prediction efficiency. The PSO-LSTM model can efficiently predict water level of the plain river network during the flood season, and the addition of artificial regulatory influences such as water conservancy projects can improve the water level prediction efficiency.

Abstract:To ensure the quality of water from Yandong and Yanzidong, which is used to recharge Yilong Lake, meets the required standards, the water quality and pollution sources of Tuanshan-Yandong-Yanzidong section of the Lujiang River in Yilong Lake were investigated and analyzed. Based on the current water pollution status in this river section, an in-situ river management system combining comprehensive interception, absorption, and purification was developed and experimentally validated. The results show that the river section is severely polluted, with total nitrogen (TN) levels significantly exceeding standards, prominent karst pollution, and water quality classified as class IV or lower. The in-situ river management system can significantly improve TN pollution in the river section, the TN contents in the overlying and interstitial water of the simulated river channel show decreasing trends, with the actual average reduction rate for TN in the overlying water rate of 49.99%.

Abstract:In order to ensure the water quality of drainage ditches flowing into the Yellow River to reach the standard and achieve ecological protection and high-quality development of the Yellow River Basin, the water quality index-development (WQI-DET) and principal component analysis index (PCAI) of water quality were used to analyze the spatiotemporal change characteristics of water quality from 2020 to 2022 in the Shizuishan section of the Third and Fifth Drainage Ditches flowing into the Yellow River. Mantel tests and multiple comparisons were employed to explore the driving factors influencing water quality in the drainage ditches, considering both natural and human factors. The results show that the compliance rate for water quality in the Helan section of the drainage ditches is the lowest (57.6% in the Third Drainage Ditch and 69.4% in the Fifth Drainage Ditch). Pollution sources are numerous in the Third Drainage Ditch, making it difficult to manage, while pollution sources are fewer in the Fifth Drainage Ditch, leading to more stable water quality. Irrigation water pollution mainly occurs in the early period of irrigation, while agricultural runoff during the mid-irrigation and winter irrigation periods can have a purifying effect on the river. Many indexes of the Fifth Drainage Ditch vary significantly between the early period of irrigation and other periods ( p ＜0.05). Natural factors have no significant effect on the overall water quality of the drainage ditches, and the local pollution sources are mainly domestic sewage and industrial discharge.

Abstract:Aiming at the problems of mutual influence between indexes, unreasonable assignment of subjective and objective weights and fuzziness of membership interval in operation safety evaluation of earth-rockfill dams, an improved combination weighting-cloud model for operation safety evaluation of earth-rock dams was proposed. Taking an earth-rockfill dam in the operation period as an object, the safety evaluation index system of the earth-rockfill dam was constructed. The subjective and objective weights of each evaluation index were calculated by the G1 method and the improved entropy weight method. The game theory was used to solve the optimal weight combination coefficient, and the cloud model theory was introduced to reduce the fuzziness of the membership interval in the evaluation process. The safety state of an earth-rockfill dam was evaluated and compared with those obtained from two traditional combination weighting methods. The results show that the evaluation results of three evaluation methods are similar and consistent with the actual situation. However, the improved combination weighting-cloud model considers the correlation between evaluation indexes, and the use of game theory makes the subjective and objective weight assignment more reasonable, which can more objectively reflect the actual operation state of the earth-rockfill dam.

Abstract:Considering the characteristics of nonlinearity and complexity of concrete dam deformation monitoring data, in order to improve the accuracy of concrete dam deformation prediction, a concrete dam deformation prediction model based on the improved empirical modal decomposition (EMD) method and the long short-term memory (LSTM) neural network was proposed. This model adopts the wavelet threshold denoising method to optimize the high-frequency components decomposed by the EMD method, effectively removing the data noise while retaining the characteristic information of the original data as much as possible. The LSTM neural network was used to perform time series prediction on the processed data. The results of case validations show that this model can accurately simulate the deformation process of the dam body, demonstrating a high prediction accuracy.

Abstract:Existing deformation prediction models for concrete dams, which rely on classical linear regression methods or shallow machine learning techniques, have significant shortcomings in extracting complex features from environmental factors and in learning the long-term dependencies of deformation-environmental factor relationships. To address this issue, this paper proposes a deformation prediction model based on the Inception module and attention mechanism-enhanced gated recurrent unit (GRU). The proposed model effectively combines the feature extraction capabilities of the Inception module with the long-term dependency learning capabilities of GRU, enabling it to extract features from monitoring sequences of dam environmental factors across different scales and to predict the long-term deformation of the dam. Additionally, by incorporating the attention mechanism, the model reduces the risk of overfitting when learning features from multiple environmental factors. Validation results from an extra-high concrete double-curved arch dam project demonstrate that the proposed model outperforms other common shallow and deep learning models at typical monitoring points, making it suitable for concrete dam deformation prediction.

Abstract:To tackle the difficulty of detecting underwater cracks in concrete dams effectively with deep learning algorithms, a semantic segmentation model for underwater cracks in concrete dams based on MobileNetV2-DeepLabv3+ was proposed. The lightweight network MobileNetV2 was introduced into the model and the deep feature downsampling multiplier was reduced to 8, so as to improve the recognition accuracy and inference speed under small dataset conditions. To alleviate the problem of category imbalance, the combination of cross entropy loss function and Dice loss function was used as the loss function of the model. The validation results from an engineering case show that the mean pixel accuracy and mean intersection over union of the model on the test set are as high as 90.87% and 86.33%, respectively, meeting the requirements for underwater cracks of high-precision semantic segmentation. Compared with other models, the proposed model shows better segmentation effect of underwater cracks of concrete dams under typical conditions, with strong generalization ability. With features of low memory usage and high inference speed, the proposed model is suitable for underwater crack detection of concrete dams.

Abstract:Aiming at the problems of the current real-time risk rate models for extra-high arch dams, such as over-reliance on monitoring statistics, difficulty in linking the physical characteristics of the structure and the critical state, and ambiguous mapping relationship between the measured data and the critical state, a real-time risk rate model for extra-high arch dams based on the finite element forward and inverse analysis is proposed. This model uses the sparrow search algorithm (SSA) and multi-output support vector regression (MSVR) inversion to determine the integrated physical and mechanical parameters of the dam body and foundation. Using the plastic strain energy criterion, it calculates the safety factor against overload of the dam in service under the limit state, identifies the critical deformation values, and derives the real-time risk rate based on the theory of engineering reliability and the safety specification of arch dams. The results of a case study of an extra-high arch dam in Southwest China show that the model can effectively link the relationship between the deformation monitoring data of the extra-high arch dam and the critical state of the project, improve the inversion efficiency of the elastic modulus of the dam body and the deformation modulus of the dam foundation, and realize the structural safety monitoring and early warning of extra-high arch dams. It can be used for long-term performance assessment and real-time risk rate control of extra-high arch dams.