When the general segmentation model is applied to the apparent cracks in the dam face concrete, the network suffers the problem of depth increasing that leads to excessive model parameters and certain loss of effective crack features. To reduce network memory occupation and feature loss, this paper develops a lightweight crack segmentation method based on a convolutional neural network. The network adopts an encoding-decoding structure, and uses a depth-separable convolution module and a lightweight feature extraction module to construct a cascade encoder; it is equipped with a decoder to fuse cross-scale information in the second stage of the encoder and to reconstruct the pixel-level geometric information lost in feature extraction to improve the accuracy of network segmentation. The experimental results show the model size of the network trained on the crack dataset of dam face concrete is 10.8 MB or a size reduction of 90.8% from U-Net, with its PA of 73.3% and IoU of 85.4%. The results verify the network is feasible in dam face crack segmentation and useful for improving the efficiency of dam face detection and maintenance.

Cascade hydropower development disrupts the continuity of natural river water temperature, leading to a significant cumulative effect on the temperature and a series of ecological effects. This study examines the water temperature along the lower Jinsha River and in the Three Gorges reservoir area , and reveals its spatial and temporal variations on different scales before and after dam construction and the cumulative effects. The impact of water temperature changes on fish spawning is also discussed. The results show that after the Xiluodu and Xiangjiaba dams were constructed, the annual mean water temperature difference along the lower Jinsha becomes smaller; along the river, the annual highest temperature at the hydrological stations shows a decreasing trend, while the annual lowest is significantly elevated, especially in January and December. After the construction, the annual temperature variation is reduced, and the time period featuring water temperature distribution has a trend of ‘convergence’－the days of water temperature distribution changed from M-type to V-type. An examination on the cumulative effect indicator finds that after dam construction, a significant time crowding effect occurs－a lag time in the extreme water temperature, water temperature delayed up to 1 - 2 months, the fish of 48% and 44% affected to a high degree at the Panzhihua section and the section downstream of Xiangjiaba dam respectively－thereby affecting severely the fish spawning and reproduction in the downstream and the fish protection sections. The study demonstrates the cumulative effects of water temperature impacted by cascade dam construction and its impact on fish spawning, laying a basis to enhance the role of temperature changes caused by large-scale cascade dams and the downstream ecological restoration.

Climate change is closely related to energy-power supply and demand. In the summer of 2022, the upper reaches of the Yangtze River experienced three worst cases: the highest temperature, longest continuous hot days, and lowest rainfall in the same period in history. This resulted in a daily power shortage of 17 million kW and 370 million kW?h in Sichuan, a major hydropower province, and imposed a significant impact on its social and economic development and people’s livelihood. To guarantee energy and power safety, it is of great significance to establish power planning mechanism and some countermeasures for power supply guarantee in extreme weather. This paper presents an analysis on the impacts of extreme weather in the 2022 summer on the power supply guarantee in Sichuan, and examines the shortcomings of previous electric power development. We suggest certain countermeasures for the period of power transformation－such as water-wind-solar-thermal energy complementarity, and a coordinated development of power supply and power grid.

Sequential power generation simulations play a critical role in the capacity configuration of hydroelectric-thermal-wind-photovoltaic-storage multi-energy complementary systems. In practice, 8760-hour system operation are hard to simulate directly using an optimization model because of its large scale. In this paper, a new time-scale decomposing technique is developed to solve this problem and realize the accurate simulations of 8760-hour system operation. Based on this, a two-stage optimization model is constructed for capacity configuration of a grid-connected multi-energy complementary system that comprises thermal power, hydropower, wind, photovoltaic, pumped-storage, and electrochemical energy storage. This new model was applied to the Qinghai power grid and achieved an optimized configuration of the system’s capacity of wind, photovoltaic, and pumped-storage.

The deformation and stability analysis of high steep slopes is a key technical problem in the construction and operation of hydropower projects. Over the past two decades, China has built a large number of large-scale water conservancy and hydropower projects. Many key technical problems of high dams and large reservoirs have been solved successfully, and remarkable progress has been achieved in the life cycle performance evolution and safety control of high, steep slopes of reservoirs. This paper takes the performance evaluation of high steep slopes in the southwest hydropower projects as the main research line, and focuses on the deformation and stability evolution of high steep slopes. We examine the research progress in determining the influencing factors of stability and failure modes of high steep slopes, stability evaluation and deformation analysis methods, seepage analysis, and safety control. The latest researches are discussed in detail on the strict three-dimensional limit equilibrium method, modified Hoek-Bray wedge method, rigid body spring method, parameter inversion method based on monitoring data, and slope seepage analysis. We also discuss the academic thinking and technical route and certain future challenges to the life-cycle deformation and stability evolution analysis of high steep slopes in hydropower engineering.

The hybrid pumped storage-wind-photovoltaic multi-energy complementary system has broad application prospects. However, its capacity design needs to characterize the complex relationship between the water volume and electric power, and its economic evaluation should consider the rules of electricity markets. This paper describes a new two-stage optimization framework for optimizing operation and capacity decision. First, a consistent assumption for the target gross output is presented; and a double-objective operation optimization model is developed. Then, a discrete decision space is obtained through optimization based on a large number of medium and long-term operation cases. Finally, the scheme with the maximized net present value (NPV) is selected. Application in a case study of the clean energy base in the upper Yellow River gives the conclusion as follows. New energy capacities corresponding to high, medium and low acceptance degrees of load loss risks are 3.2-3.9 times, 2.4-3.0 times, and 1.6-2.1 times that of the mixed pumping and storage capacity, respectively. The peak to valley ratios of the system's monthly electricity delivery range from 1.36 to 1.45, indicating the power sources in the system are well complementary on the medium and long time scales.

To systematically analyze the variations in the Qinghai Lake stage under climate change and the related influence mechanism, this paper presents a quantitative analysis of the spatiotemporal changes in the hydrological and meteorological factors of the lake basin, using a combination of time series analysis, a geostatistical interpolation method, and correlation analysis based on the data collected at the observation stations over the basin during 1956-2020. We use the global climate model to estimate the varying trends of precipitation and lake stage under different scenarios. The results show that during the past 50 years, the stage and surface area of the lake presented a general trend of decreasing first and then increasing, while an increasing trend occurred in runoff, precipitation and temperature, and a decreasing trend in evaporation. The lake stage was affected significantly by runoff, precipitation, temperature and evaporation; the correlation coefficients were as high as 0.95, 0.92, 0.88 and 0.81, respectively. We project that from 2021 to 2040, the lake stage will take a rising trend, showing an annual average rising rates of 0.218, 0.187, 0.125 and 0.132 m/a under four different scenarios. Quantitative analysis of the lake behaviors under climate change in this study lays a basis for future studies on the water cycling mechanism of plateau lakes.

Based on the fifth generation atmospheric reanalysis dataset ERA5 of the European Centre for Medium Range Weather Forecasts (ECMWF), we define two types of hot droughts or compound drought-hot events－simultaneous occurrences of meteorological drought/agricultural drought and high temperature. We examine the evolution of such an event occurred in the Yangtze River basin in the summer of 2022, and evaluate the variations in its several characteristics such as duration and spatial coverage. The results show that this hot drought began in June, became most severe in August, and weakened in September; its spatial scale varied significantly, starting from the middle and lower reaches, gradually expanding to the whole basin, and reducing to the middle and lower reaches by September. And compared with typical events in historical periods, its characteristic values were the largest. We find a significant increase in the characteristic values of the two types of compound drought-hot events in July and August from 1979 to 2022. The results deepen our understanding of the hot droughts and extreme events in a river basin and can be useful for coping with extremes under global warming.

Sediment erosion is one of the main reasons for the failure of hydraulic components of a Pelton turbine. Especially in sediment season, sediment concentration increases greatly and the process of sand-carrying three-phase flow on the surface of the runner bucket is complicated, so the erosion problem is usually more severe. The Pelton turbine at the Zixia hydropower station in Tibet, planned for construction, is examined in the present work as a case study. This paper focuses on numerical simulations of its flow process of sediment particles on the bucket surface, using the Mansouri erosion model and the Euler-Lagrange method, based on its measured sediment characteristics data provided by the design institute. The characteristics of sediment particle flow and erosion distribution over the bucket surface under high sediment concentration in sediment season are examined. The numerical predictions of surface erosion distribution agree well with previous measurements in literature. The predictions show the bucket erosion concentrates near the water outlet edge of the root and the cutout edge of the bucket. The highest similarity occurs between the distributions of sediment erosion and the impact number, so the impact number imposes a great effect on sediment erosion.

Feature extraction and intelligent recognition of the vibration signals of pump turbines are significant to reliable and safe operation of a pumped storage power station. Due to its complicated operational conditions, a pump turbine in operation can create a large number of physical sources that excite its vibrations, and the frequency components of the vibration signals are quite complicated. The traditional methods suffer a poor accuracy of feature extraction from a complicated vibration signal. To improve the accuracy, this paper describes a new model of feature extraction and intelligent recognition of the vibration signals, based on the variational mode decomposition (VMD), bubble entropy (BE), and long short-term memory (LSTM) neural network. First, this method analyzes the vibration signal using VMD and obtains several modes. Then for each mode, its BE value is calculated and a BE eigenvector is constructed. Finally, the eigenvectors of the vibration signal are trained and recognized using a LSTM neural network. We have verified the method against the complicated vibration signals measured at the top cover of a pump turbine at the Pushihe pumped storage station, and achieved a signal recognition accuracy of 97.87%, indicating its important engineering application value.

The increasing urban flooding threatens urban safety and sustainable development, and further improvement in the accuracy of urban flood risk simulation and assessment are essential to protect people's lives and properties. In this study, a flood simulation model is developed for the Shenzhen River basin based on a Cellular Automata flood simulation model, and inundation depths are simulated for the design rainfall conditions of this basin. Using the Hazard-Vulnerability risk assessment framework, we consider its different rainfall scenarios and assign weights to its different urban flood risks, through a game theory-based combination of weights determined by the analytic hierarchy process and the Criteria Importance Though Intercrieria Correlation methods. Then, its urban flood risks are evaluated and predicted. The results show typical temporal and spatial differences occur over the high-risk areas of urban flooding in the basin. With an increasing rainfall recurrence period, the ratio of the risky area increases, such as from 0.20% of a 2-year rainfall event to 0.82% of a 100-year event; the ratio of the medium-low risk area decreases, such as from 67.4% to 64.7% in terms of the two events. The results of this study can be visualized to show the distribution of flood risks over the basin, helping improve its flood control system and enhance the urban flood prevention capability and resilience of the city.

The efficiency of water energy utilization of a conventional fixed-speed small hydropower unit is often very low when its operation deviates from the design conditions. However, by examining the comprehensive characteristic curves of the turbine, a law that for a unit under the optimal guide vane opening and a certain optimal unit speed achieves the highest efficiency of water energy utilization has been found. And, a variable-speed control strategy was proposed to achieve such an efficiency, particularly for the condition of significant water head variations. This paper presents a mathematical model for simulations of a variable-speed small hydropower unit, incorporating the operation of the turbine, governor, generator and converter. A closed-loop variable-speed control system based on this new strategy was designed. Numerical simulation results indicate that compared to the fixed-speed operation scheme, the new scheme achieves a 2.39% increase in power output at the water head of 28 m. Field tests at the Zangtanqiao hydropower station verify that the small hydropower unit is successful in operating under its optimal guide vane opening and optimal rotational speed, and that the new control strategy improves power generation efficiency across high, medium, and low head conditions. The new strategy is easy to apply in the industry of small hydropower and of great significance to the real hydropower projects.

This paper focuses on the key issue of the Three Gorges hydropower station’s in-plant economic operation, which is aimed at achieving a real-time load allocation of large-scale units for minimizing water consumption. Dynamic programming usually encounters the curse of dimensionality when dealing with a large-scale hydropower unit cluster, and therefore, it cannot meet the requirement of real-time dispatching decision for the station. For training a multi-period unit load distribution model and its decision-making, we develop a deep reinforcement learning-based framework to train the deep neural network and generates unit load distribution plans through a pre-trained network model. We apply a group theory idea to processing the state and action features of the learning, so as to compress the state and action space significantly and improve model training efficiency. The results indicate that compared to dynamic programming, our new method shortens the decision-making time by two orders of magnitude at a cost of less than 1% benefit loss. Thus, it offers a rapid and efficient solution for the unit load allocations in large-scale hydropower stations.

Compared with a single reservoir, cascade reservoirs operation features a state space increasing exponentially. This paper describes a Deep Q-network (DQN) algorithm for deep reinforcement learning to solve the dimension disaster problem that is faced by the table-based reinforcement learning method in optimizing the long-term operation of cascade reservoirs. First, we derive a joint distribution function of stochastic inflow runoffs of the reservoirs based on the Copula function. Then, following the idea of time series difference, we construct a target neural network and a main neural network for approximating the values of the current action state and the next action state, respectively, and use ε-greedy algorithm to obtain optimal operation policy. Finally, the main parameters of reservoir operation are optimized by step to ensure operation efficiency. Compared with the Q-learning algorithm or its modification, the DQN algorithm improves the objective value of optimal scheduling, accelerates convergence, and avoids dimension disaster effectively in the optimization of cascade reservoirs operation.

Unsafe behaviors of construction workers are the key factor leading to safety problems in the construction process of water conservancy project. Most construction sites adopt on-site safety inspection, wearable equipment, real-time monitoring, and other methods to identify unsafe behaviors, but such methods are time-consuming, laborious, expensive with low information level, and unfavorable to the timely discovery and early warning of dangerous behaviors. This paper presents a new method of unsafe behavior identification suitable for large scenes of water conservancy projects based on computer vision and deep learning. First, an improved method of YOLOv5 is developed to solve the problem of missing and wrong detection of small objects in construction workers and machinery in large scenes, and a multi-object object detection model of construction workers and machinery is constructed. Then, based on the target detection model, recognition methods are suggested for each of the routine unsafe behaviors, such as being close to the static danger area, dynamic construction machinery, and not wearing safety helmet, etc. Engineering application verifies that our identification method strengthens the means and intensity of construction site control and improves effectively the level of water conservancy engineering construction safety and intelligent control.

A deterministic model of reservoir optimal operation is usually difficult to be directly applied to practical operation. To improve applicability, we develop a new joint optimal operation model for the Yalong River cascade reservoirs and use the Gaussian radial basis function (RBF) to extract operation rules, and then calibrate the rule and verify its fitting errors. The results show the annual power generation of the cascade is increased by 98.32 billion kW?h or an increase of 8.73% compared with its design value. The new operation processes optimized by the model for the Lianghekou, Jinping I and Ertan reservoirs follow certain common patterns. The determination coefficient of Gaussian RBFs is larger than 90%, and it can fit the operation rules effectively. This achieves a simulated power generation greater than the design value by 51.27 billion kW?h (+5.67%).

How to generate a slip surface is one of the key issues in slope stability analysis, and to construct a slip surface with a complex shape is quite challenging. The curve based on the integral of a Logistic function or other basis functions can take the place of the traditional slip surface generated using non-circular curves, which not only improves the smoothness of the slip surface but reduces its degree of freedom. Slip surface generating methods based on basis functions are sorted out in this paper. They are divided into the superposition method of continuous functions and the integral operation method. The characteristics of the Logistic function and its integral are examined carefully, and the influence of different parameters on the curve shape is obtained to narrow parameter ranges. This provides a basis to reduce the degree of freedom of slip surfaces. And a method is presented to determine the initial parameter values of a given slip surface. The calculations of two slope examples show our method is tolerably good in slope stability analysis.

Hydraulic concrete is prone to erosion and cracking in the wet-dry alternating zones under water level fluctuation. A mixture of aerobic and facultative anaerobic mineralized microorganisms can better adapt to the fluctuating oxygen content in concrete cracks. This study prepares a microbial self-healing agents using the microorganisms of aerobic Bacillus megaterium and facultative anaerobic Sporosarcina pasteurii for two cases－a single component and a mixture of both, and conducts the compressive strength test and permeability coefficient test on the concrete samples. For each case, we evaluate the influence on the performance and crack self-healing effect of hydraulic concrete through a quantitative index analysis of crack repair. An optimal mixing ratio of aerobic and facultative anaerobic has been obtained. A SEM analysis of the samples is used to reveal the microscopic mechanism of mixed microorganisms on concrete performance improvement. The results show that the mixing of mineralized microorganisms can effectively elevate the density of concrete pore structure, and the best effect of calcium carbonate precipitation occurs at the mix ratio 4:6 of Bacillus megaterium and Sporosarcina pasteurii. Under the synergistic influence of aerobic microbial respiration and facultative anaerobic microbial enzymatic action, the mixed microorganisms yield better improvement on the mechanical properties and crack self-healing efficiency of hydraulic concrete.

To study the durability of hydraulic concrete under the environment of freeze-thaw salt intrusion in the northwest region, we prepare concrete specimens with different fly ash dosages and conduct freeze-thaw cycling tests, using different concentrations of sodium sulfate solution as the medium. The tests clarify the specimens’ behaviors under different cycles－appearance, quality, compressive strength, and dynamic modulus of specimens. And a concrete lifespan prediction model is developed based on the XGBoost model, and it is evaluated and validated. The results indicate that as the number of freeze-thaw cycles increases, the quality, compressive strength, and dynamic modulus of concrete gradually decrease; The number of freeze-thaw cycles and the concentration of sodium sulfate solution are the key factors of concrete lifespan. The 8% solution causes the highest degree of damage, and the corresponding rate of concrete quality loss reaches 4.55% after 150 freeze-thaw cycles. The fly ash content has a certain impact on concrete durability; its optimal value is 10% and the resulted quality loss rate is 3.99% after 150 freeze-thaw cycles. The results show the XGBoost model has high accuracy and reliability in predicting concrete lifespan. This study would help the durability design and lifespan predictions of concrete structures.

For the time series prediction issue of dam deformation, a spatiotemporal multi-dimensional input matrix of deformation is derived considering the correlation of deformation at multiple measuring points; an Informer-AD dam deformation prediction model is constructed that integrates multi-dimensional spatiotemporal information based on K-means clustering. We use the K-means clustering to partition rationally the deformation measuring points, then apply a panel data regression model to integrate the analysis of spatiotemporal dimensions and partition results. Finally, we develop an Informer-AD dam deformation prediction model to integrate multi-dimensional spatiotemporal information. This model is used to learn spatial feature sequences and integrate spatial features through a fully connected layer to output predicted dam deformation values. Its application to a concrete gravity dam shows that our prediction method, considering spatiotemporal correlation, can fully explore the relationship of the overall state of dam deformation versus the spatial distribution characteristics of measuring points. It better captures the spatiotemporal characteristics of deformation values and thus improves prediction accuracy, which implies that our model has a high accuracy and satisfactory applicability, useful for engineering application.