In order to study the applicability and error sources of reasoning formula method in design flood calculation in complex hilly areas, Morris screening method was used to analyze the sensitivity of reasoning formula method parameters. The applicability of reasoning formula method in different watershed areas was analyzed by comparing with hydrological comparison method and historical flood survey results. The results show that the area average rainfall is the main error source of the reasoning formula method, and the error tends to increase with the area of the basin. The reasoning formula method can still be applied to the basin with an area of 500-1 000 km² by reasonable selection of parameters. Taking Pujiang River Basin as an example, using the traditional area rainfall calculation method (without reduction), when the control section catchment area is about 100 km², 100-500 km² and more than 500 km², the error is 5.1%-5.3%, 11.2%-16.0% and 25.2%-25.5%, respectively. When the basin area is larger than 500 km², the calculated result is only about 0.2% smaller than the historical survey data after reasonable reduction of area rainfall.

In order to explore the spatial and temporal distribution of water quality in the Yellow River basin and the causes of pollution, the monthly monitoring data of 7 basic water quality indicators at 8 provincial boundary stations of the Yellow River from 2015 to 2019 were systematically collected, and the spatial and temporal distribution characteristics and trend changes of water quality of the Yellow River were systematically analyzed by seasonal Kendall trend test method. Based on the surface water quality standard, the status of water quality was evaluated and the main pollution indicators were determined. The linear regression model was used to discuss the causes of pollution. The results show that the spatial distribution characteristics of most water quality indicators are middle reaches > lower reaches > upper reaches. The concentration values of water quality indicators at Tongguan station are significantly higher than those at other stations. TP concentration values at some stations in the middle and lower reaches of the river show a significant upward trend, especially at Gaocun and Lijin stations. From the perspective of spatial variability, the poor water quality of Tongguan station is related to the convergence of several seriously polluted tributaries in this section. From the perspective of seasonal variability, the seasonal distribution characteristics of TP concentration show that the non-flood season is less than the flood season. Under the effect of rainfall erosion in the flood season, agricultural phosphorus carried by tributaries flow into the river with the surface runoff formed by rainfall, resulting in the increase of TP concentration. The seasonal variation of NH3-N concentration in flood season is less than that in non-flood season, which is strongly correlated with air temperature. Higher flow rate and air temperature in flood season contribute to the dilution and degradation of NH3-N.

In order to efficiently utilize water resources in the upper reaches of the Minjiang River Basin, based on hydro-meteorological data of the basin from 1960 to 2018, a variety of statistical theories were used to analyze the trend, cycle and abrupt point of runoff. The elastic coefficient method based on Budyko theory (Choudhury-Yang formula) and the comparative method of cumulated slope change rate (SCRCQ method) were used to analyze the attribution of runoff changes. The results show that the runoff decreases significantly at a rate of 12.5 mm/10a, and a sudden change occurred in 1993. There are 3 years, 8 years, 12 years and 18 years main cycles of runoff change, and the annual distribution is uneven, mainly concentrated in summer and autumn, accounting for 73.6%. Rainfall is the most sensitive to runoff, followed by underlying surface parameters, and potential evapotranspiration is the least sensitive. Choudhury-Yang formula and SCRCQ method calculated that the contribution rate of climate change to runoff reduction was 58.13% and 63.29%, respectively. Climate change was the main factor of runoff reduction in the study area, and the contribution rate of human activities to runoff reduction should not be underestimated, reaching 41.87% and 36.71%, respectively.

In order to review the power generation capacity of the Three Gorges Reservoir under normal operation conditions, and explore the impact of different dispatch modes and upstream reservoir regulation, the preliminary design scheme, optimized operation scheme, and 2015/2019 edition scheme were taken as the calculation model. Considering the main control reservoirs in the upstream, the natural runoff series in Yichang station from 1946 to 1975 used in the preliminary design were selected to calculate the natural flow derivation for reflecting the inflow runoff characteristics of the Three Gorges Reservoir in the current stage. Using natural/return runoff as four dispatch models’ scenarios, the power generation capacity was analyzed. The results show that under current stage, the power generation capacity of the Three Gorges Reservoir is 100.8 billion kW·h, 14.3% higher than the design value; By optimizing the dispatch mode, the full storage rate in October can be improved. Compared with the design mode, the power generation capacity is increased by 5.5-9 billion kW·h; By the regulation of the upstream reservoir, the inflow runoff and the utilization rate of water resources is improved. The power generation capacity is increased by 0.8-2.8 billion kW·h.

Phasor measurement unit (PMU) can measure phasors directly, which provides an important technology for dynamic safety monitoring. Aiming at PMU measurements experience errors, a dynamic state estimation method for generators based on cubature Kalman filter (CKF) was proposed. In this method, the fourth order dynamic equations of generators were taken as the system function. The fading factor was introduced into CKF to keep the residual sequence orthogonal at all times, which improves the adaptability of the estimation algorithm. The adaptability of the estimation algorithm was improved, and the defect that the estimation result deviates from the real value due to the uncertain parameters of the generator model was overcome. Simulation results verified the effectiveness of the algorithm.

It is significant to achieve safe operation of the project and strengthen the construction of the whole life cycle safety supervision system of the dam to scientifically and rationally conduct a comprehensive evaluation of the earth and rock dam seepage control treatment scheme. From the four guideline layers of engineering risk removal effect, engineering management, engineering economy and engineering construction, this study constructed the evaluation index system of earth and rock dam seepage control treatment scheme. Aiming at the evaluation of the risk attitude of decision makers in the process of seepage control program selection, the risk attitude factor was introduced to assign interval evaluation to the indicators, and the subjective and objective weights were optimized based on game theory. And then the program was analyzed and evaluated using the improved TOPSIS model. The effectiveness of the preferred system was tested by taking a reservoir anti-seepage treatment alternative as an example. The results show that the method is feasible and effective for the selection of seepage control solutions.

In order to evaluate the matching degree between water conservancy project for tobacco and tobacco drought in Guizhou Province, the daily rainfall data from 17 meteorological stations, tobacco production and water supply data of water conservancy project for tobacco during 2006-2017 was used to construct a matching degree evaluation model, and grade standard of matching degree was proposed. The results show that the water conservancy project for tobacco alleviated tobacco drought to a certain extent. However, the matching relationship between water conservancy project for tobacco and tobacco drought still needs to be improved. The matching degrees of Qianxinan, Qiannan, Anshun, Liupanshui, and Guiyang were 55.4%, 52.1%, 46.2%, 45.9%, 44.7%, and 34.4%, respectively, which were in the fourth grade with a poor matching degree. The matching degrees of Bijie, Qiandongnan, Tongren, and Zunyi were 31.7%, 28.6%, and 26.8%, respectively, which were in the fifth grade with a worse matching degree. In different growth stages, water shortage in root extension period mainly occurred in 90% and 95% effective rainfall frequency, and water shortage in prosperous and mature periods mainly occurred in each effective rainfall frequency. Additionally, the lowest matching degree occurred at the mature period. It is suggested that in the future, the layout of the water conservancy project for tobacco should focus on strengthening the construction in Zunyi, Tongren and Qiandongnan, and improve the rationality of the layout of water conservancy project for tobacco, as well as enhance the guarantee ability of drought resistance of tobacco.

As a common safety problem in water diversion engineering, water hammer is often caused by abnormal operation of valves, So, it is necessary to conduct a comparative analysis of water hammer problems under different valve adjustment methods. The CFD technology and dynamic mesh technology were used to simulate different single and double valve adjustment modes, and analyze the fluctuation of water hammer pressure at different locations in the pipeline. The results show that when the single valve is adjusted, the first fast and then slow closing valve mode has an inhibitory effect on the water hammer within a certain range of fast shutdown; The fast closing amount is too large, which will play a worsening effect of water hammer; When the fast shutdown amount is 60%, the water hammer pressure can be effectively controlled, and the water hammer will deteriorate when the fast shutdown amount is 80%; When adjusting the double valve, considering the pressure before and between the two valves, it can be found that the “first 0.25 s off 60%, the last 0.25 s off 40%” method can avoid the occurrence of excessive pressure and negative pressure. The research result provides a design and operation scheme for the rapid shut-off process of the valve in practical engineering.

In order to solve the current situation that hydraulic turbine fault diagnosis relies excessively on expert experience and has low efficiency, the historical data of hydraulic turbine fault and relevant expert experience were used to establish the fault tree model for seeking the mapping between risk hidden danger and fault diagnosis. Through the conversion of fault tree model and Bayesian network model, the probability importance and sensitivity of root node were deeply studied by using the reverse diagnosis technology of Bayesian network. Hydraulic turbine components and fault causes were inferred to realize the fault diagnosis.

In order to solve the cavitation damage problem of the spillway with high head, three schemes were proposed. The aeration tank was added at pile number 0+130.73, 0+161.78 and 0+186.39, respectively, and the corresponding Fr was 6.65, 7.91 and 8.53. By means of RNG κ-ε turbulence model and physical model test, the aeration concentration and cavitation number indexes of the section prone to cavitation erosion were analyzed. The results show that the effect of adding aerator at pile 0+186.39 is better than the other two schemes, and the bottom aeration concentration is above 15% and the flow cavitation number is greater than 0.3 in the section prone to cavitation erosion, the cavitation failure problem in the drainage channel section can be effectively solved.