With the acceleration of urbanization in China, the demand for water resources is increasing, and the leakage of the water supply pipe network has caused waste of water resources. Therefore, leakage control is imminent. The first is to accurately calculate the leakage of the pipe network, but the current methods have certain deficiencies. Based on the Kalman filter, this paper establishes an analysis model of pipeline network leakage, and uses the empirical formula of node leakage, standard Kalman filter (SKF) and innovation adaptive Kalman filter (IAKF) to calculate the leakage of the pipeline network, respectively. Taking the actual pipe network project as an example, the EPANET simulation was carried out, and the calculation results were compared. The results show that using both filtering algorithms is more accurate than the nodal leakage formula, and the result of the IAKF algorithm is better than that of the SKF algorithm. This provides a theoretical basis for accelerating the research on leakage control technology and implementing scientific and effective leakage control methods, which has certain significance for urban water saving and sustainable development.

The paper aims to comprehensively grasp the carrying capacity of water and soil resources in Ningxia, and promotes the coordinated development of water and soil resources in arid and semi-arid areas. An evaluation index system was constructed based on DPSIR model. The CRITIC-entropy weight-TOPSIS was applied to evaluate the carrying capacity of water and soil resources in Ningxia from 2012 to 2019. The obstacle degree model and R/S method were used to explore the influencing factors and predict the future trend. The results showed that from 2012 to 2019, the carrying capacity of water and soil resources in the whole Ningxia region shows an upward trend, but the overall situation is basically overloaded. The carrying capacity in Guyuan City is greatest in the Ningxia region. Based on the obstacle analysis, the driving force subsystem has the strongest restriction on the carrying capacity of water and soil resources in Ningxia. Economic density, natural population growth rate, per capita GDP, grain output per unit of arable land and per capita water resources are the main factors affecting the water and soil resources carrying capacity in Ningxia. According to the R/S method, in the future, the water and soil resources carrying capacity in Ningxia will show an upward trend. Guyuan and Zhongwei have a significant upward trend, while other cities have a weak upward trend.

Due to the uncertainty of runoff, there are certain risks when formulating hydropower station power generation plans. In this paper, the concept of spectral risk measurement in economics was introduced in reservoir dispatch to quantify risk, and a benefit-risk equilibrium optimization model was established, which can reflect the attitude of decision makers to risk through different risk aversion coefficients and confidence values according to the actual situation of the power station. Taking the Three Gorges Power Station as an example, the results show that compared with the traditional model of maximum power generation benefit, the proposed benefit-risk equilibrium optimization model can effectively reduce the risk caused by the uncertainty of inflow in hydropower stations under the condition that the loss of power generation income is extremely small.

Aiming at the problem of urban waterlogging caused by heavy rainstorms in coastal cities, this paper takes Fuzhou City as an example and constructs a fined coupling model of urban one-dimensional drainage network and two-dimensional land surface hydrodynamics based on Infoworks ICM. The model was applied to the analysis of rainstorm waterlogging in Gutian Road area of Fuzhou during Typhoon "Lubi" in August 2021, which verified the high simulation accuracy of the model. Based on this model, the rainstorm waterlogging process in Gutian area under different recurrence periods was analyzed, and two "sponge" reconstruction schemes were proposed. It is found that increasing the pipe diameter is more conducive to improving the waterlogging situation in the study area. The study can be used to accurately calculate the process of urban underground drainage and land area water in Fuzhou City, which provide technical support for the real-time decision of flood control and drainage in coastal cities.

After the operation of the Three Gorges Project, the process of water and sediment in the lower reaches of the reservoir changed greatly and led to the adjustment of the evolution law of the curved channel, which may have adverse effects on the stability of the river and the safety of flood control. Based on the measured data, the evolution law of the typical bends in the lower reaches of the reservoir and its main causes were analyzed. The results show that after the operation of the Three Gorges Project, the typical bends in the lower reaches of the reservoir all appear "cutting beaches", and the main reason is that the sediments enter each river reaches. The amount of sand, especially fine-grained sand, has been greatly reduced. The main reasons for the "skipping bends" in most of the bends in the Lower Jingjiang are that the variation coefficient of the monthly average runoff process is large and the bending radius of the bends itself is small. With the successive operation of cascade reservoirs in the main and tributaries of the upper reaches of the Yangtze River, it is expected that the variation coefficient of the monthly average runoff process in the lower reaches of the reservoir will continue to increase, which may lead to the occurrence of bends with relatively small bending radius that do not currently have "skipping bends".

Dispatching operation of hydropower station is in face of the large capacity and large units, unit control new scheduling features such as irregular more limits and more complicated operation requirements. To aim at power station safe, stable and economic operation, a model of economic operation inside the station was established to realize optimal load distribution of hydropower station units under various operation conditions. The optimal output range of each unit under different head was analyzed. The optimal load distribution rule of the power station was studied, and the economic operation model of the station was established. Based on the economic operation model of the station, the optimal load distribution table was designed. The optimal load distribution result table of the power station under different loads, different heads, and different unit commitment was proposed to guide the real-time economic scheduling of the power station. Taking Hongjiadu station as an example, the results show that the model can reasonably arrange the start-stop sequence and number of units according to the real-time load instruction, and give real-time start-stop suggestions of units. It can obtain the optimal load distribution combination of the proposed unit under various working conditions, reduce the monthly average water consumption rate by 0.18%, reduce the adverse working conditions such as frequent start-stop and frequent crossing of the vibration zone by 9.5%, ensure the efficiency of the unit operation, and improve the economic benefits of the power station.

Under the background of global climate change, river flood risk assessment is of great significance to disaster prevention and mitigation in river basins. By analyzing the spatiotemporal change characteristics of floods in the Nanliu River Basin from 1960 to 2021, a flood risk assessment index system was established for flood risk assessment. The results show that the interannual variation of flood disaster in Nanliu River shows a fluctuating upward trend; 1993 was a big catastrophe year of flood disaster change; On the whole, the upstream basin, the middle river valley and the downstream estuary plain are the flood prone areas, and the flood risk shows a distribution pattern of high in the southeast and low in the northwest, which is consistent with the characteristics of historical flood data. The results can provide scientific basis for the comprehensive management of Nanliu River Basin.

The cluster forecasting model is a long-term forecasting physical model based on mathematical statistics. This method has been applied in the long-term forecast of the upper reaches of the Yangtze River. After years of practice, there is space for improvement in accuracy and efficiency of the model. This paper selected the upper reaches of the Yangtze River as the demonstration area for forecast, the monthly areal rainfall from June to August as the forecast object, and used the coincidence rate of anomaly symbol and the percentage of anomaly error as the test methods of the model. And then the construction process and improvement of the model were described. Finally, the improved model was used to forecast the monthly area rainfall in the upper reaches of the Yangtze River from June to August during 2020 to 2022. The prediction results show that the improved model has a great improvement in the anomaly sign agreement rate and the percentage of anomaly error, and it has been implemented easily, which has certain reference value.

Early and timely defect detection is essential to ensure the safe operation of hydraulic concrete structures. The deep learning-based computer vision method does not require complex manual feature engineering, and can automatically determine the category of structural defects in remote images, overcoming the shortcomings of traditional manual vision that are labor-intensive, subjective and prone to errors. Inspired by this, this paper proposes a deep learning-based defect detection method, which introduces attention mechanism into the ResNeXt50 network to adaptively recalibrate the channel-level feature responses, so that the model can pay more attention to the defect information in the image and enhance the feature extraction ability. Test results show that the proposed method can achieve an F1 score of 88.0%, and realize a good classification effect for common concrete defects.

The effect of luring fish is one of the key factors affecting fish passing efficiency in the entrance area of fish-passing facilities. Proposed a new optimization idea of side-channel drop-flow to lure fish. The upstream water was replenished to the area outside the fishway entrance as drop-flow. With the help of auxiliary facilities such as cantilever plate, the attractive flow was formed in the horizontal orientation. Take the fishway project of Shaping I Hydropower Station as example, a large scale physical model was established to demonstrate and optimize the scheme, and the flow pattern near the entrance of the fish-passing facility were analyzed, and the possibility of attracting fish was discussed. The experimental results show that the one-dop scheme is favorable for attracting the migratory fish in the horizontal orientation, with the flow velocity of 0.3-0.5 m/s, and the information of flow pattern is rich in the entrance area, which is suitable for fish aggregation.