In natural river channels， aquatic vegetation typically grows in the areas near the riverbanks， which in turn affects the stability of the riverbank and the discharge capacity of the river channel. During the growth of vegetation， the inundation degree （ratio of water depth to vegetation height， H/hv） will change with the change of vegetation height or water level， which will affect the flow characteristics.By constructing a three-dimensional numerical model of an open channel with submerged vegetation groups distributed along the river bank， the Large Eddy Simulation （LES） method was used to study the average flow field and turbulent structure around the vegetation group under four submergence levels （4.0、3.0、2.0 and 1.5）. The results indicate that， when submerged vegetation is distributed on one side of the open channel， the maximum cross-sectional velocity deviates from the channel center， appearing on the non-vegetation area， and gradually approaches the water surface as the submergence ratio decreases. The presence of submerged vegetation generates a mixing layer in the vertical direction， intensifying flow turbulence. High turbulent kinetic energy regions are distributed along the interface between the vegetation and non-vegetation area as well as at the vertical interface between the submerged vegetation and the upper water layer. As the submergence ratio decreases， the size of the horizontal shear layer vortices increases. When the submergence level is 1.5， the horizontal shear layer vortex can extend into the non-vegetated area. At the same time， the intensity of the secondary flow in the cross-section strengthens， and the turbulence within the entire flow field becomes more intense.

Exploring the escape swimming behavior of fish under non-steady flow not only helps to understand the ecological adaptability and survival strategies of fish， but also provides scientific guidance for fish resource conservation and habitat restoration projects under hydropower development. Based on the fish escape swimming behavior experiment and numerical simulation in a compound open channel， the relationship between the fish distribution area and spatial distribution of hydraulic factors is analyzed， the fish preference for different hydraulic factors is revealed， and the critical hydraulic conditions for fish escape swimming behavior are identified. The results indicate that the most common area where experimental fish appear is the slope of the channel-shoal junction. The residence coefficient is negatively correlated with the rate of flow change， while exactly opposite in the main channel area. The preference coefficient of experimental fish for flow velocity， turbulent kinetic energy， integration time scale， and spatial velocity gradient is roughly similar， showing a “bell” shaped distribution trend that increases first and then decreases， and there is an obvious preferred suitable interval. The characteristics and time response of fish channel-shoal migration behavior change with different water rise and fall durations. The channel-shoal migration behavior of fish varies in response to different duration. The key hydraulic factors that trigger fish migration behavior under rising and falling conditions correspond to the average flow velocity difference and down-ramping rate， with critical thresholds of 1.25~1.34 BL/s and 1.96~3.08×10^-2 BW/s， respectively.

As a typical industrial pollutant-receiving lake， North Lake located in Qingshan District （Chemical Industry Zone） of Wuhan City has long been subjected to persistent water pollution and ecological degradation. Clarifying the evolutionary patterns of water quality deterioration and their driving factors is critical for implementing targeted remediation strategies. Based on the water quality monitoring data from 2014 to 2023， this study systematically evaluated the spatiotemporal evolution characteristics and key pollution factors of North Lake using the trophic level index （TLI）， comprehensive pollution index （CPI）， and rank correlation coefficient （RCC） methods. The results showed that： ① From 2014 to 2020， the water quality of North Lake exhibited an overall deterioration trend， with annual average classification declining from Class IV to inferior Class V. The CPI increased from 0.54 to 1.50， while the eutrophication level escalated from mild to moderate （TLI peaking at 65）. Key pollutants—ammonia nitrogen （NH?-N）， total nitrogen （TN）， and total phosphorus （TP）—reached peak concentrations of 4.05、6.72 and 0.336 mg/L， respectively， in 2018. ② A significant improvement occurred in 2021， with annual water quality stabilizing at Class V and the proportion of months with inferior Class V water quality dropping from 100% in 2017 to zero. By 2023， NH?-N， TN， and TP concentrations decreased to 0.33、1.31 and 0.138 mg/L， representing reductions of 91.85%， 80.51%， and 58.93%， respectively. ③ RCC analysis demonstrated notable pollutant reduction trends for BOD?， COD， and TN （ r_s = -1.00） during 2019–2023. However， the pollution load ratio of TP increased annually， reaching 26.71% in 2023， and the TLI remained at moderate eutrophication levels （≥60）， highlighting the need for sustained TP control and ecological restoration. This study innovatively integrates multi-period dynamic analysis and multi-method comprehensive evaluation， unveiling a three-phase evolution pattern of water quality （deterioration–transition–stabilization） in North Lake. It identifies key pollutant factors and provides a scientific foundation for long-term monitoring and management of water pollution in industrial lakes across the middle and lower reaches of the Yangtze River.

Meandering rivers exist widely in nature， and understanding their complex evolution process and mechanisms is important. Many studies have been conducted to investigate the movement of water and sediment， as well as the evolution of riverbeds in meandering rivers， using river model tests. This paper reviews the research results of existing meandering river models and systematically summarizes the development process， characteristics， and application scope of idealized river models and natural river models. The main results are as follows： the idealized river models make a high degree of generalization of natural rivers. It can be categorized into three types of meandering river models based on different design approaches： arc and straight line combination， sine-derived curve， and Kinoshita-derived curve. These idealized river models are more applicable in fundamental theoretical studies of river sediment dynamics. The natural river model is based on an actual river and is designed according to a certain similarity principle. It mainly includes the scale model and the natural evolution model. The scale model is a physical model of the prototype river scaled according to a certain geometric scale that ensures the similarity of water movement and sediment movement. The scale model is suitable for the inversion and simulation of water and sediment transport， channel siltation and degradation， and riverbed evolution of the prototype river channel. The natural evolution model is a model river similar to the prototype in riverbed morphology， which is shaped by controlling the flow conditions and river boundary conditions such as riverbed composition and riverbed gradient. The natural evolution model is more applicable in studies related to the evolution of river types. Both types of natural river models bear similarities to the prototype river and can provide a direct reference for practical engineering applications. In summary， idealized river model and natural river model have their own advantages and disadvantages in basic and applied research， and different model types can be selected according to the research needs.

Hydrocyclone is a commonly used solid-liquid separation equipment， widely used in fields such as agricultural irrigation， water conservancy engineering， and tunnel inflow for water and sediment separation treatment. Due to the high-speed rotation of the fluid， a high turbulence zone is formed inside the hydraulic cyclone， which can easily cause secondary flow phenomena. The secondary flow is opposite to the main direction of motion， which not only consumes energy， but also causes collisions between particles， changes the trajectory of particle movement， and increases the number of fine particles trapped in the overflow and bottom flow. In order to solve the problems of short-circuit flow and overflow of coarse particles in traditional hydrocyclones， a new type of hydrocyclones with integrated filter screen was proposed in this stydy. The Euler Lagrange liquid-solid multiphase turbulence model based on particle dynamics theory was applied， and the CFD-DEM coupled numerical simulation method was used to simulate the flow of five different particle sizes in the solid-liquid two-phase flow inside a traditional hydraulic cyclone and a new filter type cyclone. The velocity field， turbulent field and particle motion and force in the two hydrocyclones were compared and analyzed. And the separation efficiency was used as the evaluation index to verify the rationality of the simulation. The simulation results show that 60 μm particles have the longest average residence time inside the traditional cyclone and the new filter cyclone， 120 μm particles have the shortest average residence time inside the traditional cyclone， and 180 μm particles have the shortest average residence time inside the new filter cyclone. This indicates that the addition of the filter prolongs the residence time of coarse particles inside the cyclone， making separation more complete. The larger the particle size， the greater the fluctuation amplitude of the average resultant force acting on the particles. When the particles first enter the cyclone field， the trend of the average resultant force acting on the particles in the traditional cyclone is relatively gentle， while the trend of the average resultant force acting on the particles in the new filter cyclone is more severe. This indicates that the added filter has played a certain promoting role in the force acting on the particles. Compared with the traditional hydrocyclone， the maximum average tangential velocity in column section and cone section is increased by 57.6% and 22% at five different inlet velocities. The maximum upward average axial velocity decreased by 14.6%， the maximum downward average axial velocity increased by 11%， the average residence time of particles increased by 1%~6.9%， and the recovery quality of bottom flow port of particles increased by 4%~87%. The average tangential contact force curve of particles fluctuates sharply first and then gently， which can reach the state of stable separation in advance.

To ensure the safety and stability of the siphon system during the water filling phase， this study investigates the dynamic response of the unnel-type inverted siphon structure under conditions where water is already present. Using the Yushi Factory siphon from the Central Yunnan Water Diversion Project as a case study， we employed computational fluid dynamics （CFD） software for numerical simulations， complemented by physical model experiments. This analysis focused on the flow velocity， pressure， and stabilization time of the siphon structure under various water filling flow rates， starting from an initial saturated state. The results indicate that the trend of head loss changes initially decreases with increasing flow rate， followed by an increase. This finding aligns with the change trend observed in the model test data， and the discrepancies between the two are minimal， suggesting a high level of accuracy in the numerical simulations. Furthermore， the inertial time constant of the water flow shows significant numerical consistency with the predicted stabilization time from the numerical simulations. As the flow rate increases， the stabilization time of the water flow exhibits a decreasing trend， providing crucial reference data for the operational procedures of filling the siphon structure.

Compared with the gorge type reservoir， the planar morphology of the plain sandy river reservoir is more susceptible to incoming water and sand conditions， so it is of great significance for reservoir management to study the morphological evolution of the plain sandy river reservoir. In this paper， a typical plain and sandy river reservoir， Xixiayuan Reservoir on the Yellow River， is selected to study the evolution law of reservoir plane morphology. Xixiayuan Reservoir has been in a state of continuous siltation since its construction， and the flow path of the reservoir has been constantly evolving， which has profoundly changed the morphological characteristics of the reservoir. This paper analyzes the morphological evolution of Xixiayuan reservoir area and its driving factors based on the measured cross-sectional topographic data and the image data of Xixiayuan Reservoir area from 2007 to 2023. The results show that reservoir area has transformed from multiple flow paths to a single flow path， the height difference between the beach and the channel decreases， the curvature and branching coefficient of the channel decrease significantly after 2020， and the swing width of the deep pool swings greatly in the lower section of the reservoir， especially near the dam section. The changes of the morphology of the reservoir area are affected by many factors， such as the conditions of inflow of water and sediment and the operation of the reservoir. Among them， the swing width of the xxy05 section is significantly positively correlated with the average annual inflow of inflow of water and sediment in the previous three years， the number of days when the annual average inflow of water is greater than 4 000 m3/s， and the number of days when the average daily water level before the dam is 130~131 m in flood season. Before 2020， the longitudinal stability of the mainstream road in the broad reach of Xixiayuan Reservoir showed an overall increasing trend， while the lateral stability decreased. After 2020， the longitudinal stability decreased， and the lateral stability was basically stable. The evolution process of reservoir morphology in plain sandy river reservoirs is a prerequisite for the development of silt reduction measures， and the process of incoming water and sand should be paid attention to in time during the scheduling process， and based on the safety of the project， the development of reasonable sand discharge scheduling strategy. This paper provides a case study for the morphological evolution of the plain sandy river reservoir area， and provides suggestions for the operation and dispatching of the sandy river reservoir.

In traditional fishway design， the rationality of the fishway design is generally verified through model tests due to the influence of many factors such as flow velocity， flow pattern， terrain and longitudinal slope of the river channel in the project area. Considering that the three-dimensional numerical simulation of the flow field has advantages such as flexibility， simplicity and more complete result data compared with physical model tests， in the fishway design work of SPL Hydropower Station Project， after fully studying the migration habits of upstream fish， the full terrain three-dimensional numerical simulation was selected to numerically simulate the flow field in the tailwater area of the hydropower station. After analyzing the flow pattern of the river channel in the project area， it was proposed to set a straight tailwater channel after the reverse slope of the tailwater of the power station， and to arrange the fishway inlet on the left or right bank of the tailwater channel. The recommended scheme was verified by three-dimensional numerical simulation calculation. The results show that the flow field of the fishway scheme arranged on the left bank of the straight tailwater channel is basically smooth， and the flow pattern at the fishway inlet position meets the requirements， which is more in line with the actual conditions of this project. The method of arranging fishways in this paper that based on the full terrain three-dimensional numerical simulation flow field， which can provide new ideas and successful case references for the design and layout of fishways in the future.

Revealing the coupling coordination relationship between water resources utilization and high-quality economic development in the Yellow River Basin is very important for improving the macroeconomic regulation of natural resources and economy society， as well as promoting regional sustainable development. Taking cities along the Yellow River in Henan as the research object， an index system of high-quality economic development and water resources utilization was established in this study. Based on the panel data of Henan Province from 2012 to 2022， the comprehensive empowerment-TOPSIS method， which combines entropy weight method and coefficient of variation method， is used to measure the water resources utilization and high-quality economic development level；Combined with the coupling coordination degree model and the relative development degree model， the coupling coordination development trend and interaction mechanism of the two are analyzed. The results show that： ① During the study period， the water resources utilization efficiency of the cities along the Yellow River in Henan fluctuated and increased. The water resources utilization efficiency of Zhengzhou city was the highest， and the water resources utilization effect of Xinxiang city was the worst； ② The overall economic quality development index of cities along the Yellow River in Henan has been continuously improved， with Zhengzhou having the highest economic quality development index and Puyang having the lowest economic quality development level； ③ The cities along the Yellow River in Henan Province as a whole showed the characteristics of high quality economic development synchronized with water resource utilization， but there were obvious differences among cities； ④ The coupling coordination degree between water resources use efficiency and high-quality economic development in cities along the Yellow River in Henan Province is on the rise， but the level is still low. In terms of time series， it has undergone the evolution process of forced coordination - primary coordination， and the spatial distribution pattern is south of the Yellow River > north of the Yellow River， and the coupling coordination evolution process of each city is different. It is necessary to combine the characteristics of cities， pay attention to energy conservation and environmental protection， improve water-saving measures in agriculture and industry， and strengthen scientific and technological innovation capabilities， so as to improve the coupling and coordination of water resources utilization and high-quality economic development.

Groundwater is an important water supply source in Jinan City. In order to achieve sustainable development and utilization of groundwater resources， it is necessary to carry out groundwater storage changes analysis and groundwater resources assessment in Jinan. This paper is based on the GRACE gravity satellite and GLDAS global land surface assimilation system data， and uses singular spectrum analysis method to reconstruct terrestrial water storage change （ΔTWS） data. A multiscale geographical weighted regression （MGWR） model was constructed， using evapotranspiration （ET）， precipitation （PRE）， land surface temperature （LST）， and vegetation index （NDVI） as regression variables to downscale groundwater storage change （ΔGWS） data in Jinan， and the spatiotemporal change analysis of ΔGWS and groundwater resource evaluation were conducted. From April 2002 to December 2023， ΔTWS and ΔGWS showed a downward trend in Jinan. During the year， ΔGWS was the smallest in June and the largest in December. The lag effect of ΔGWS on precipitation changes is obvious. The MGWR model enhanced the resolution of ΔGWS data in Jinan from 0.25° to 1km. The trend changes and spatial distribution of ΔGWS exhibited a high degree of consistency before and after downscaling， indicating that the MGWR model is effectively applicable for downscaling ΔGWS data.The BEAST algorithm is used to deconstruct ΔGWS time series in Jinan. ΔGWS is mainly controlled by the trend term， and there are two mutation points with a probability greater than 0.5. There are obvious differences in the spatial distribution of ΔGWS in Jinan. Laiwu District and Gangcheng District have relatively large groundwater storages， while Pingyin County and Changqing District have lower groundwater storage. The ΔGWS center of gravity is mainly located in Licheng District. The water amount in the rising section of ΔGWS is used to characterize the amount of groundwater resources in Jinan. From 2003 to 2019， the average amount of groundwater resources is 2.290 billion^{m3， and the amount of groundwater resources in extraordinary dry years， dry years， and normal flow years are 1.657、2.722 and 2.955 billionm3 respectively in Jinan. The research results provide scientific reference for the rational development and utilization of groundwater resource in Jinan.}

In response to the current situation in traditional unsaturated infiltration theory， where the determination of the diffusion coefficient D（θ） involves complex equipment， numerous experiments， and poor universality， this study， under the first-type infiltration boundary condition， calculates the wetting front position at time t through Laplace transform. It simultaneously solves the Boltzmann transform and variational transform， deducing explicit expressions for the diffusion coefficient， water content， and infiltration distance within the wetting front. The calculation results indicate that， when the Laplace transform path assumes a constant diffusion coefficient， the Boltzmann transform path shows that the diffusion coefficient is a quadratic function of distance and a cubic function of moisture content. For the variational transform path， the diffusion coefficient is a linear function of distance. All regression equations have passed the L2 norm residual test， confirming the accuracy and validity of the results. A comparison between the measured and theoretical values of the diffusion coefficient shows that the maximum relative error between the measured diffusion coefficient from horizontal soil column experiments and the theoretical value calculated using the Boltzmann transform path is 8.2%. The maximum relative error between the measured diffusion coefficient and the theoretical value calculated using the variational transform path is 7.7%. Both errors are below 10%， meeting the engineering requirements. The joint calculation method proposed in this paper relies solely on the mathematical background and path relationships of classical transformations， without introducing additional assumptions or requiring field experiments. This approach provides a new reference for calculating unsaturated infiltration parameters.

Recent global climate and land use changes have increasingly attracted public attention， and their impact on the global hydrological cycle has become a hot topic for discussion and research among governments， society， and researchers. This paper focuses on the typical watershed of the NiuLan river-Dianchi reservoir replenishment project in Kunming City as the research area. Based on collected multi-period climate and land use data from the study area， the climate change trend is analyzed using climate tendency rates and the Mann-Kendall （M-K） trend test method. The historical and future spatial-temporal changes in land use are analyzed using the CA-Markov model and GIS spatial overlay tools. Furthermore， the SWAT model is employed to simulate watershed hydrological processes under different scenarios and investigate the response of watershed runoff to climate and land use changes. The study finds that： ① From 1965 to 2014， the temperature in the study area rose at an overall rate of 0.256°C/10 years， while precipitation decreased at a rate of 18.5 mm/10 years. The M-K mutation test indicates a significant rising trend in average temperature in the watershed after 1993， while there were multiple mutations in annual precipitation during this period， but the changes were not significant. ② Cultivated land， forest land， shrubland， and grassland are the main types of land use in the study area. Since 2005， land use changes have primarily manifested as a continuous expansion of construction land， a decrease in grassland area， and a slight reduction in cultivated land area， resulting in an increase in water yield. ③ When compared with the contributions of land use changes， climate change is found to be the primary driving force behind changes in runoff in the study area， with precipitation having the most significant impact on runoff. Under the combined effects of climate and land use changes， runoff during the flood season is particularly sensitive； under the action of the two separately， land use changes lead to increased runoff during the flood season and decreased runoff during the dry season. Variations in monthly runoff under climate change are significant， with changes during the flood season being more pronounced compared to the dry season.

Aiming at the problem that the non-stationarity of the daily inflow runoff series of power stations is aggravated and it is difficult to predict it accurately， a prediction model of the daily inflow runoff series based on the Improved Whale Optimization Algorithm （IWOA） with multiple mechanisms is proposed. Firstly， the Variational Mode Decomposition （VMD） is used to project the inflow runoff data into sub-sequences with different frequencies. Then， the quantum bit Bloch coordinate coding strategy is added to the traditional whale optimization algorithm to improve the quality of the initial solution. At the same time， the nonlinear decreasing strategy of the convergence factor and the sparrow warning mechanism are introduced to improve the global optimization ability of the algorithm. The IWOA is used to optimize the hyperparameters （ maximum number of hidden nodes， maximum number of candidate nodes ） of the Stochastic Configuration Network （SCN）， and then the SCN is used to predict the decomposed modal components. The VMD-IWOA-SCN combined prediction model is constructed to improve the prediction accuracy. The Particle Swarm Optimization （PSO）， WOA and the proposed IWOA are selected to verify the optimization performance through the CEC2019 test function. The comprehensive analysis shows that the improved IWOA algorithm with multiple mechanisms has the best optimization performance. A total of 4018 daily inflow runoff time series data of a power station from 2011 to 2021 were selected for case analysis. Compared with other popular optimization algorithm models， root mean square error （RMSE）， mean absolute error （MAE）， mean absolute percentage error （MAPE） and Nash-Sutcliffe Efficiency （NSE） were selected to evaluate the prediction results. The results show that the NSE of the VMD-IWOA-SCN combined model reaches 99.09 % in the prediction experiment， which is significantly higher than other models. It is shown that the proposed VMD-IWOA-SCN prediction model can provide more accurate and stable daily inflow prediction results， which can provide effective reference for hydropower station power generation prediction， reservoir scheduling and other decision-making.

Parameters calibration is an important part of improving the runoff simulation accuracy of hydrological models. Constructing efficient and robust parameter optimization algorithms is a hot topic in applied research on hydrological modelling. To solve the premature and easy-convergence problem in the butterfly optimization algorithm， this paper introduced the Halton sequence to improve the homogeneity of the initial population， balanced the global and local search process by using a dynamic transformation probability， and improved the search equations by adopting the Lévy flight and standardized operation operators. An modified version of butterfly optimization algorithm （MSBOA） was established and applied to the parameter optimization of the Xin′anjiang model. The convergence and stability of the original Butterfly Optimization Algorithm （BOA）， the modified Butterfly Optimization Algorithm （MSBOA）， the Sine Cosine Optimization Algorithm （SCA） and the Whale Optimization Algorithm （WOA） in the parameter calibration of the Xin′anjiang model were compared and analyzed. The application results in the upper Qinjiang River basin above the Luwu hydrological station showed that： ① compared with the original algorithm BOA， MSBOA yielded an enhanced Nash-Sutcliffe efficiency coefficient of 0.03， a lower relative bias within ±5%， and a reduced root mean square error by 8.2%. This indicated that parameters calibrated by the MSBOA had a better performance in daily streamflow simulation， which can accurately describe the watershed hydrological process. ② All four algorithms had good applications in the parameters calibration of the Xin′anjiang model （Nash-Sutcliffe efficiency coefficient was higher than 0.85）. In terms of convergence， MSBOA had faster convergence speed and the highest Nash-Sutcliffe efficiency coefficient， followed by WOA and SCA algorithms. In terms of stability， MSBOA achieved the lowest standard deviation of Nash-Sutcliffe efficiency coefficient and the smallest fluctuation range of parameterization results. This indicated that the MSBOA yielded high stability， followed by WOA algorithm. This proposed MSBOA can enrich the parameter optimization study on hydrological model and provide a new way for model parameter optimization.

Water consumption forecasting plays a significant role in the sustainable management of water resources， enhancement of production efficiency， and protection of ecological environments. To improve the precision of water consumption forecasting， this paper proposes an Improved Grey Wolf Optimization BP neural network （IGWO-BP） model for water consumption prediction. A total of 16 influencing factors were selected， and four factors with significant correlations to water consumption were identified through grey relational analysis， serving as input variables for the IGWO-BP model， with water consumption as the output variable. The Grey Wolf Optimization （GWO） algorithm was enhanced by incorporating Circle chaotic mapping， a nonlinear convergence factor， Levy flights， and a greedy search strategy. The performance of the IGWO algorithm was compared with that of the traditional Grey Wolf Optimization （GWO）， Genetic Algorithm （GA）， and Particle Swarm Optimization （PSO） algorithms through six basic function tests to validate the optimization and convergence capabilities. Subsequently， the IGWO-BP model was applied to forecast the water consumption of Binzhou City， and its predictive results were compared with those of the GWO-BP， PSO-BP， and BP neural network models. The results reveal that among the 16 influencing factors selected， total population， agricultural land， average temperature， and annual sunshine duration are most closely related to water consumption. Under identical parameter settings， the IGWO algorithm significantly outperformed the other three algorithms in terms of optimization and convergence in the performance tests of six basic functions. In water consumption forecasting， the IGWO-BP model demonstrated superior convergence speed and a lower minimum fitness value compared to both the GWO-BP and PSO-BP models. Furthermore， the IGWO-BP model exhibited a substantial reduction in average relative error and root mean square error compared to the other three models. Whether evaluated based on model fitness values or prediction accuracy， the IGWO-BP model delivered enhanced water consumption forecasting performance and higher prediction precision， providing a novel approach and technical support for future water consumption predictions.

Intelligent grouting technology plays a critical role in hydraulic engineering and foundation reinforcement. Currently， grouting pressure and unit grouting rate are primarily adjusted manually， leading to low grouting precision and frequent overpressure incidents. This study designs an intelligent grouting pressure control system， integrating a pressure regulation platform， intelligent control unit， and upper-level master control center. The system is capable of controlling pressure and processing real-time data， ensuring the stability and safety of the grouting process. The system's hardware includes key components such as a grouting pump， diverter valve， inflow meter， pressure gauge， outflow meter， and electric control valve. By integrating a PLC with the electric control valve， real-time adjustments to the return pressure are made， ensuring stable and precise pressure control. A data acquisition and pressure stability control logic program for the grouting process are developed， and an integral sliding mode controller is applied for pressure regulation. The controller's performance under dynamic and disturbance conditions is verified， demonstrating its ability to respond quickly to pressure changes and adapt to different grouting requirements. This enhances the automation and intelligence of the grouting operation. This provides a solid foundation for the widespread application of intelligent grouting technology and offers significant potential for future use. This work presents a comprehensive design of an intelligent grouting system， explores the effects of various pressure control algorithms， and confirms the system's performance in real-world applications. The research results show that the intelligent grouting system operates stably and has precise pressure control， which provides a solid foundation for the widespread application of intelligent grouting technology and has broad application prospects.

In response to the problems of inaccurate thermal parameters， poor reference of empirical values， low accuracy of traditional back analysis methods， and high dependence of BP neural network on initial weights and thresholds in finite element simulation of thin shell concrete prestressed aqueducts， this paper constructs a neural network with four input layers， five hidden layers， and three output layers according to actual engineering requirements. The Northern Goshawk Optimization algorithm （NGO） is used to optimize the BP neural network， and the NGO-BP neural network for thermal parameter inversion of concrete prestressed aqueducts is simulated and trained using finite element simulation samples. It is applied to the thermal parameter inversion of a bidirectional prestressed U-shaped aqueduct， and the calculation results of on-site monitoring data， traditional BP neural network method， and BP neural network method optimized by Genetic Algorithm （GA） are compared. The results show that the NGO-BP neural network significantly improves both computational accuracy and convergence efficiency. The finite element results obtained from parameter inversion are in good agreement with monitoring data， which can better guide the construction practice of concrete prestressed aqueducts.

The construction of shipping hubs may face complex road conditions such as multiple bends and slopes on the access road. To explore the stability of the transportation of large gate equipment under these complex road conditions， an gate transport behavior simulation model was established using a large domestic shipping hub as an example. The model was used to analyze the transportation stability of shipping hub gates under various access road conditions. The results show that the stability of gate transportation decreases with the increase of vehicle speed， and the radius and slope of the entrance road have a significant impact on transportation stability. For curved road sections， the lateral load transfer rate |LTR| value should be controlled within 0.43 when the large gate transport vehicle is driving at low speed （v ≤ 20 km/h）， and within 0.53 when driving at high speed （20 km/h < v ≤ 30 km/h）. For slope sections， in order to prevent vehicle braking failure and rollover accidents， it is necessary to reasonably match the transportation speed and slope. When the transportation vehicle speed is 20 and 30 km/h， the slope should be controlled within 7.2% and 6.3%， respectively. The research results can provide reference for the scientific formulation of transportation plans for shipping hub gates.

Timely identification of accident information， such as accident type， accident location， accident occurrence time and accident flow rate， is a necessary condition for the long-term and stable operation of the long distance water transmission pond， and is also an inevitable requirement for the construction of water conservancy intelligence. Based on the hydraulic response model under the accident state， this paper proposes an accident identification model of canal and tunnel system， which includes three parts： accident type judgment and alarm， accident location and accident time calculation， and accident flow process calculation. The accident identification model is tested on the HEC-RAS model of the main trunk canal of Guumutang. The results show that： ① The model can effectively identify the accident condition， and no misjudgment will be made in the canal without accident； ② The accident identification rate of the model is 95.8%， and the false alarm rate is 0%， which basically meets the requirements of accident identification. ③ The minimum accident flow identified by the accident identification model is 0.15 m³/s， which can meet the needs of practical applications.

The Gongzui Power Station has been in operation for more than 50 years， and the penstock has been situated in the exposed pipe section of the facility. Recent observations indicate that wall thinning due to abrasion poses potential risks for safe operation. In view of the above problems， stress testing， along with digital strength and modulus analyses of the penstock subjected to hydraulic turbine load shedding， has been performed， and a stress and strain monitoring system for the penstock has been established. This monitoring system possesses real-time monitoring capabilities and an alarm forecasting feature. Currently， the wall thickness of the penstock has diminished to approximately 28 mm， and research indicates that there will be no incidents of tube explosion or similar events during load rejection. However， if the wall thickness decreases to about 24 mm， the maximum equivalent stress of the penstock during load rejection will approach the safety threshold. It is advised that the maintenance or replacement of the pipeline system should be strengthened after the wall thickness of the steel pipe is reduced to around 24 mm due to wear and other factors.

Efficient and accurate detection and analysis of concrete dam defects images is necessary to ensure the safe operation of the dam. Aiming at the problems of multi-category and uniqueness of surface defect images of concrete dams and the lack of high-quality samples for on-site inspection and inspection of existing dams， this paper first simulates typical defects such as cracks and leakage through self-made concrete slabs to construct a multi-category defect data set of concrete dams， and then uses Yolov8 + U-Net “two-step method” to establish a multi-category surface defect detection and analysis model of concrete dams. Finally， the surface defect images of on-site inspection and inspection of a concrete gravity dam is taken as the test object， and the established detection and analysis model is used for intelligent detection. The results show that the “two-step method” model based on Yolov8 + U-Net algorithm can realize the efficient and accurate detection of seepage and crack defects in concrete dams. The recognition and positioning accuracy rate of the model is 0.84， the recall rate is 0.98， the segmentation accuracy rate is 0.91， and the recall rate is 0.71.

Siphon-type termite nests， with their complex structures and hydraulic characteristics， have a complicated impact mechanism on dike seepage stability and the potential safety hazards are very serious. This study analyzes the destructive mechanisms by which siphon-type termite nests， influenced by fluctuating floodwaters， can lead to dike collapse and breaches during their dynamic development. Taking a typical earth embankment in the Dongting Lake area as the research object， a stress model of the upper soil of the siphon ant tunnel was constructed. And conditions for maintaining stability during high water levels and flood recession were derived based on the principle of limit equilibrium analysis. Using finite element software， the influence of siphon-type termite nests on the seepage characteristics and stability of the dike was analyzed， and recommendations for termite control were proposed based on the research results. The findings indicate that as the diameter of the siphon-type termite nest increases， there is a significant rise in the phreatic line， creating a high-pressure zone around the main nest. The termite tunnels begin to conduct water only when the water level surpasses the highest point of the nest. During the early development stages of the nest， the seepage on the downstream slope increases sharply and stabilizes as the nest reaches a larger scale. The safety factor of dikes containing termite nests of different sizes decreases by 1.5% to 15%， depending on the scale of the nests.

Accurate groundwater level prediction is of great significance for groundwater resources management in irrigation areas. However， due to the difficulty of processing unstructured well position data in irrigation areas， traditional deep learning methods do not consider the spatial positional relationship of each monitoring well and the own attributes of each well in irrigation areas， resulting in a serious lack of physical interpretability of deep learning models. This study fully considers the spatial positional relationship of each monitoring well and the static attributes of well points to construct a graph structure with sufficient physical meaning. A GWN model based on graph neural networks is proposed. Taking the Yichang irrigation area in the Hetao irrigation area as an example， combined with the EEMD algorithm to preprocess the groundwater level data， the model groundwater level prediction results are compared with the two baseline models LSTM and GRU to verify the rationality of the model. The results show that the GWN model with EEMD algorithm is better than the baseline model， and the statistical indicators such as MAE， RMSE， and NSE are better than the baseline model. At the same time， compared with the single well prediction of the baseline model， the GWN model can predict the water levels of all monitoring wells at the same time， and the prediction efficiency is significantly improved. In addition， the graph structure containing physical meaning is adaptive and iteratively updated to reproduce the groundwater network with physical background in the irrigation area. The relationship and reasons between the wells can be intuitively interpreted from the adjacency matrix combined with the static attribute distribution map. In short， the model has good groundwater level prediction effect and interpretability， which can provide a reference for the groundwater level management of the Yichang irrigation area in the Hetao irrigation area of Inner Mongolia.

The Kubuqi Desert， the seventh largest in China， confronts significant challenges from recurrent agricultural droughts and salinity， primarily due to the progressive depletion of groundwater resources and excessive， imprudent irrigation practices that intensify soil salinization. Therefore， Hangjin Banner， Ordos City， Inner Mongolia Autonomous Region， was designated as the study area for this research， with the objective of developing a time-series monitoring system to assess regional soil moisture content （SMC） and sunflower growth. Based on extensive time series data from Landsat 8， MODIS， and UAV aerial photography， combining several vegetation indices， this study thoroughly assessed sunflower growth circumstances， and used different drought indices to derive regional sunflower soil moisture data. The findings indicated that： ① The Vegetation Water Stress Index （VSWI） demonstrates high accuracy （r>0.65） in capturing SMC dynamics； ② The efficacy of drought indices in reflecting SMC dynamics in sunflowers is influenced by factors such as variable characteristics， temperature， and crop cover， resulting in diminished performance of VSWI， TCI （Temperature Condition Index）， VCI （Vegetation Condition Index）， VHI （Vegetation Health Index）， and TVDI （Temperature-Vegetation Dryness Index）； ③ The integration of VSWI with the Normalized Difference Red Edge （NDRE） and the Green Normalized Difference Vegetation Index （GNDVI） facilitates real-time monitoring of crop water requirements and nutritional status. The techniques presented in this work effectively extract the growth status and soil moisture data of sunflowers， therefore offering scientific assistance for water management and appropriate irrigation of crops in arid and semi-arid locations.

The severe shortage of water resources in Hebei Province has long relied on overdraft and overextraction of groundwater to support economic and social development， resulting in a series of ecological and environmental problems such as the drop of groundwater level， river cut-off and land subsidence. The current agricultural irrigation water quota， promulgated in 2017， can no longer meet current and future water resource management needs. Therefore， taking Jinzhou City of Hebei Province as an example， this study adopts the theoretical value method to adjust the irrigation quota of Jinzhou City from the perspective of water saving and emission reduction. By analyzing the precipitation frequency from 1960 to 2020 in Jinzhou City， the effective rainfall under different hydrological year types was calculated. The daily reference crop evapotranspiration during the entire growth period was calculated by Penman formula， and the irrigation quotas of winter wheat and summer corn under different hydrological years were calculated by crop water requirement， effective precipitation， groundwater infiltration recharge and non-engineering measures of water saving. The results showed that the adjusted irrigation quotas in wet， normal and dry years were 111.6、309.24 and 364.67 mm respectively， laying a good foundation for water resources management and social sustainable development in Jinzhou City.

In order to meet the demand for water rights allocation and trading in water-scarce areas， this study proposed an evaluation method for the water-saving capacity of canal lining in large and medium-sized irrigation districts. Taking an irrigation area in the northwest of China as an example， the theoretical calculation method and the field measurement method were used to assess the amounts of theoretical and actual water-saving， and to evaluate and analyze the water-saving capacity of an irrigation district water-saving renovation project in Northwest China from 2018 to 2020. The results indicate that the amounts of theoretical and actual engineering water-saving in lined canal sections were 42.87 million m3 and 45.08 million m3， respectively， and when converted to the canal head， the amounts of theoretical and actual engineering water-saving were 47.18 million m3 and 48.61million m3， showing close agreement. The research process and results can provide a technical basis for the assessment of water-saving capacity， and offer technical support for promoting groundwater extraction control and water right transfer in irrigation districts.

The trading of water rights in rural weir ponds is a “high-efficiency” and “high-quality” means to promote the sustainable use of water resources.Currently， while pilot practices of water rights trading for rural ponds have emerged sporadically across China， these explorations are primarily experimental and lack systematic， broadly applicable trading models. This limitation hinders the market-oriented promotion of pond water rights trading. This study selects five provinces—Hunan， Hubei， Anhui， Sichuan， and Jiangxi—as representative regions and focuses on the core concepts of asset-based and intensive utilization of pond water resources. It systematically explores effective pathways for market-oriented redistribution of pond water rights.Based on two scenarios—pond water rights certified to village collectives and certified to individual households—the paper proposes two corresponding trading models， with village collectives and individual households as transferor entities， respectively. The applicability of these models in different contexts is analyzed in depth. Considering the characteristics of pond water resources and actual water usage needs， the study further proposes diversified and multi-level market-oriented trading models to accommodate the varied water usage scenarios in rural areas.To promote the standardized development of the pond water rights trading market， the paper provides four recommendations concerning institutional development， infrastructure improvement， and service guarantees. These include advancing rural pond water rights system construction， enhancing rural water network infrastructure， promoting simple monitoring devices and water rights trading platforms， and encouraging the growth of third-party service agencies.The findings of this study offer significant theoretical support and practical guidance for the market-oriented trading of pond water rights in rural areas. They not only contribute to the governance of rural micro water bodies and the maintenance of small-scale water conservancy facilities but also explore new avenues for rural water resource management and market-oriented reform.

In engineering practice， the vortex rope and induced pressure pulsations in the draft tube of a Francis turbine pose significant threats to the stable operation of the unit. To address this， two draft tube structure optimization schemes were designed and the flow characteristics and pressure pulsation characteristics in the tailwater pipe were analyzed by numerical simulations. By comparison， it was found that although the pressure pulsation amplitude can be effectively reduced after installing a guide device or adding a groove， adding a guide device to the tailwater pipe is more efficient. The results suggest that incorporating guide vanes in the draft tube can enhance turbine stability and improve performance across a wide load range.

In order to study the energy loss mechanism of the overflow parts of a vertical mixed-flow pump during operation， this paper takes the unit of a pumping station as the research object， adopts SST k-ω turbulence model and entropy production theory to conduct steady numerical simulation of five flow conditions of the mixed-flow pump under the Angle of the designed blade， and deeply discusses the energy loss mechanism of the pump. The results show that the numerical simulation results of hydraulic performance of mixed-flow pump are basically consistent with the model test results， and the error is less than 10%. There is a positive correlation between the entropy production of inlet domain and runner domain and the pump flow rate， and the entropy production increases with the increase of flow rate. However， the entropy output value of the whole， guide vane domain and exit domain is the lowest near the design flow. The energy loss in the runner domain can be divided into turbulent entropy generation and wall entropy generation， and the main energy loss comes from flow separation and dynamic and static interference phenomena on the blade surface. The turbulent entropy production in the guide vane region accounts for a large proportion， which is caused by the mismatch between the inflow Angle and the guide vane relief Angle， resulting in the generation of vortex structure and the increase of entropy production.

The pressure fluctuation can induce vibration and noise in pumps， threatening the safe and stable operation of the unit. Taking a certain super-large flow and high-power large centrifugal pump as the research object， a three-dimensional full-passage model of the prototype pump was established. The numerical simulation method was adopted to study the regulating effect of the flexible guide vane on the energy characteristics， internal flow characteristics and pressure pulsation characteristics of the centrifugal pump. The results show that the flexible guide vane can significantly improve the circumferential non-uniformity of pressure in the fixed guide vanes， effectively suppress the flow separation in the fixed guide vanes， fully recover the kinetic energy by reducing the radial velocity of the fluid， and shift the dominant frequency of the pressure fluctuation in the impeller to the low-frequency direction， which is conducive to improving the efficiency and operation stability of the pump.

To investigate the super-hump limiting water level of pumping stations for the smooth start-stop transition of the pumping units under the new cut-off mode. According to the new model of fast gate cutoff with flapper， the corresponding mathematical model of super-hump start-stop transition process is established， and the mathematical model is solved by MATLAB software to obtain the characteristic curve of axial pump unit in the process of start-stop transition， and then to determine the super-hump limit level at which the axial pumping station can be operated safely. The results of the study show that in the super-hump water level startup process， compared with the traditional open gate startup mode， the new mode of closed gate slapper startup not only does not cause river water backflow but also has smaller starting power of the unit； Two-stage shutdown effectively avoids the pump accident shutdown into the flyaway condition， ensuring that the axial flow pump unit operation safety； Combined with the start/stop characteristics of the unit， it is determined that the pumping station can operate stably at a water level of 3.30 m over the hump. The conclusions of this study are expected to provide theoretical guidance for the start-stop transition process of axial flow pumping stations under over-hump water level.