In the long service process of the deep excavated expansive soil channel slope, the real-time identification of the service behavior of the channel slope by means of displacement monitoring index is an important means to ensure the safe operation of the channel slope. Therefore, it is of great significance to formulate reasonable monitoring indexes for the safety of canal slope engineering. The POT model was presented in this paper. On the basis of introduction of 2 σ criteria and chauvenet principles, the threshold T and the corresponding relation between monitoring index was established. Through the best solution to satisfy the relationship between threshold, a canal slope displacement monitoring index for improved POT model was established to apply to the Taocha section of South-to-North Water Transfer Project of deep excavation expansion earth canal in slope engineering. Compared with the traditional POT model, the results show that the improved POT model based on 2σ -chauvenet criterion can effectively avoid subjectivity and random error, and it has higher calculation accuracy. The proposed displacement monitoring index is more safety, which has stronger guiding significance to prevent the risk of canal slope and ensure the safe long-term operation of canal slope.

In order to solve the control problem of rainfall runoff and pollutants in highway service area, SWMM model was used to establish hydrology and water quality model of Tieshan service area. Based on the selection principle of LID measures and the site conditions of Tieshan service area, the control efficiency of rainfall runoff and pollutants in the service area were studied under four schemes of without LID measures, combination of infiltration pavement and green roof, combination of sunken lawn and green roof, and combination of permeable pavement and sunken lawn under different rainfall return periods. The results show that the three combined LID measures can effectively reduce the rainfall runoff, peak runoff, pollutant concentration and total pollutant load in the service area, among which the LID combined measure scheme of permeable pavement and sunken lawn has the best effect on pollution interception and emission reduction. In addition, the LID measures are more effective in controlling rainfall runoff and pollutants when the rainfall intensity is low than that is high.

Aiming at the problem that many measured thermometer data are not effectively used in the previous prediction of crack opening and closing time series data, and there are multiple correlations between their variables, considering the advantages of principal component analysis (PCA) in dealing with multidimensional data and gate recurrent unit (GRU) neural network in dealing with complex time series data, this paper constructed the PCA-PSO-GRU combined prediction model. Taking the monitoring data of the opening and closing of the left inducing joint of a concrete gravity arch dam as a sample, the principal components of the input variables were extracted to reduce the dimension of the input data. And then the model training and multi-step prediction were carried out. The mean absolute error and root mean square error were used to evaluate the prediction accuracy of the model. The prediction results were compared with PSOGRU, PCA-PSO-BP and the traditional statistical regression models. The results show that the PCA-PSO-GRU combined prediction model has higher accuracy in the prediction of inducing joint time series data, which can provide guidance for the evaluation of opening and closing degree of dam inducing joints.

For the 1# bulb tubular unit of Baisha Hydropower Station, the prestress vanishment of the rotor bracket and hot sleeve of the spindle could results the scratching of the generator and the uneven air gap between stater and rotor, which generates the excessive unilateral magnetic pull. This problem may lead to a serious accident in runaway or load rejection conditions. According to the experiential and theoretical analysis, the solution points to the fastening connection by a pin arrangement between rotor bracket and main shaft. Consequently, the unit has been running steadily under various working conditions. As there are numerous large, medium and small units in China facing the similar problem, the treatment scheme and design calculation methods proposed in this paper could be considered as a typical solution reference.

In order to solve the heat dissipation problem of heat pipe rectifier cabinet for excitation system of giant hydropower station, the working principle of loop heat pipe and the processes of development and improvement of loop heat pipe rectifier cabinet were introduced. The advantages of integrated loop heat pipe and horizontal symmetry structure, as well as its application in Wudongde Hydropower Station were analyzed. The practice shows the loop heat pipe rectifier cabinet adopting completely self-cooling heat dissipation mode is in a thermal stable state after long-term operation in the field, and a series of problems, such as dust, noise, power supply, maintenance, can be solved from the design principle. Meanwhile, it can improve the safety, reliability and automation level of the equipment and reduce the energy consumption, the benefits of energy-saving and environmental protection are obvious.

Aiming at the difference of the runoff fluctuation regime after the construction of reservoir, this paper adopted several fluctuation quantification methods to analyze the difference of downstream runoff fluctuation regime after the construction of Three Gorges reservoir from the dispersion, tendency and steepness change degrees on the basis of nearly 70-year long time series of runoff data at Yichang hydrologic station. Compared with the natural runoff regime, the results show that the dispersion degree of runoff decreases (mostly in flood control storage capacity reservation period), the tendency change degree of runoff increases and the steepness reduces (mainly in drawdown period), and totally the fluctuation regime changes intricately.

Based on the GMS software, a numerical model of groundwater flow in the Linfen Basin was established to predict the groundwater level under different driving factor scenarios, and the impact of different driving factors on groundwater level changes was quantified. The results show that only in the precipitation scenarios (dry, normal, and wet), the variation of shallow groundwater is greater than that of confined groundwater. When the precipitation is from 458.8 mm to 568.0 mm, the absolute value of the difference in the annual average water level change rate of the shallow groundwater is 0.11 m; Only under the pressure extraction scenario (extraction volume in 2018 as the benchmark), both shallow groundwater and confined groundwater have great changes. When the pressure extraction increases from 0% to 50%, the absolute value of the annual average water level change rate difference between shallow groundwater and middle-deep confined water is 0.16 m and 0.25 m, respectively; Under the situation of the rising water level of the Fenhe River, only the shallow groundwater on both sides of the river will be affected to a certain extent. The research results can provide a certain scientific basis for the rational utilization of groundwater in the Linfen Basin.

Taking Jinghe River Basin in Gansu Province as the main research object, the Jinghe River trunk canal irrigation area in the upper reaches of Jinghe River was selected as a typical experimental irrigation area. The irrigation return water, farmland water content and crop growth trend of Jinghe River trunk canal irrigation area were monitored. The water consumption of Jinghe River trunk canal irrigation area was tested by the combination of diversion and drainage difference method and multi-level soil water balance model (VSMB model). According to the monitoring experiment, the total water diversion volume of Jinghe River main canal irrigation area in 2018 is 20.76 million m³, and the total water return volume is 5.23 million m³. The water consumption coefficient of Jinghe River main canal irrigation area is 0.56 by using the diversion and drainage difference method. The total annual water diversion of typical plots in Gannong experimental field is 1 901 m³, and there is no retreating water on the surface. The water consumption coefficient obtained by diversion and discharge difference method is 0.75, and the water consumption coefficient obtained by VSMB model is 0.71. The research results provide reference for the study of water consumption coefficient of other agricultural irrigation areas in Jinghe area, and have certain guiding significance for irrigation area planning and design and high-efficiency agricultural irrigation pilot work.

The research on water level prediction method of Fenshui River control station in Fenshui River basin that improves the prediction accuracy is of great significance to support regional flood control decision-making. The typical historical floods of Fenshuijiang hydrology station from 2012 to 2020 were selected and classified into category Ⅰ and Ⅱ flood based on systematic cluster analysis method, with the maximum 1-day flood volume, peak discharge, peak water level and the proportion of flood initiation duration as cluster indexes. The water level classification and prediction of water level stations were carried out by flow routing and water level and flow relation conversion method. The upper and lower boundaries of the water level and flow relationship of different flood types were constructed, and a classified interval prediction method of flood level was proposed. The results show that the accuracy of flood peak water level prediction for the two types of floods is improved effectively, the width of forecast interval is obviously reduced, and the availability of forecast results is improved.

Huaxi Reservoir is a typical karst mountain reservoir and an important water source in Guiyang City. In order to find out the distribution of heavy metals in the sediment of Huaxi Reservoir, a total of 21 sets of sediment samples were collected from two cross sections. The content of Cr, Cu, Zn, As, Cd, Pb and other heavy metal elements were determined by inductively coupled plasma mass spectrometer (ICP-MS). Their distribution characteristics and content levels on the cross section were analyzed. The potential ecological risk assessment method and geo accumulation assessment method were used to evaluate the pollution risk, and systematically analyze the degree of heavy metal pollution in two different sections of the reservoir. The results show that the content of heavy metals in the section of inlet reservoir is low and lower than the background value, and its concentration sequence is Zn>Cr>Cu>As>Pb>Cd; The heavy metal content has a law of decreasing with the increase of the offshore distance, the main reason is caused by the influence of surrounding agricultural activities; The content of heavy metals in open sections is relatively low, and the concentration sequence is Zn>Cr>Cu>Pb>As>Cd; The heavy metal content increases from the left bank to the right bank, which is mainly related to river velocity and sediment particle size. The heavy metal pollution in the sediment of the reservoir is light pollution, and the pollution degree of the section near the reservoir is right bank>left bank>river center, and the pollution degree of each heavy metal in the open section gradually increases from left bank to right bank.