To investigate the latest variations of incoming water and sediment in the Yellow River Delta under the influences of climate change and human activities， this paper analyzes systematically the changes in water and sediment regime at Lijin Station during 1950-2018 from four aspects of water and sediment， including trend， variability， annual distribution and water sediment collocation using the series of hydrological verification methods. The results show that： ① there was a significant decrease in runoff and sediment load on the whole. ② The water and sediment regime experienced the variation in around 1985， with the timing closely associated with the construction and operation of reservoirs. ③ The annual distribution of water volume and sediment load also changed. The data shows that the water and sediment delivery to the sea mainly occurred in the period between August and October during 1950-1999 and the non-uniformity of annual distribution presented an increasing trend. However， after the year 2000， this phenomenon usually occurred in the operation period of the Water-sediment Regulation Scheme （WSRS） from June to July， and the annual non-uniformity declined slightly. ④ Before 2000， the incoming sediment coefficient increased gradually， but after 2000 it decreased dramatically to a level in natural stages of 1950-1959， and there was a decreasing trend for the flow discharge corresponding to the maximum sediment transport capacity. In summary， the runoff and sediment transport process of the YRD has gradually transferred from natural mode to man-made mode.

With the continuous improvement of urban sewage collection and treatment capacity， agricultural non-point pollution has gradually become the main pollution source of natural rivers in the Yellow River Basin， especially in the Yellow River Irrigation Area， whose treatment has become the key point and difficult point of water pollution control. At present， the projects of “rural water system improvement” and “villages with enchanting water” provide conditions for the ecological purification of the returned water in the Yellow River Irrigation Area. In this paper， the habitat characteristics， nitrogen and phosphorus removal capacity and environmental adaptability of three dominant submerged macrophytes in the Yellow River Basin are analyzed and compared， which provideds a scientific basis for the construction of distributed submerged macrophyte wetland system based on the interception during transportation and in-situ treatment of agricultural non-point pollution in the Yellow River Irrigation Area.

High precision numerical simulation of one-dimensional complex flow in natural rivers is still faced with great difficulties under the condition of rapid change of cross-section geometry. Based on the Godunov Scheme， a flux calculation method for HLLC approximate Riemann Solver based on the conservation type Saint Venant Equation is proposed. The solver is extended from the shallow water equation to the conservative Saint Venant Equation. A variable space reconstruction method for natural rivers with complex cross-section geometry is proposed. According to the equivalent principle of flow section area and static moment， the cross-section is generalized into a rectangle. The cross-section geometry at the interface of the element is constructed by linear interpolation. According to the reconstruction results of water level， the reconstruction values of flow section area and static moment on both sides of the interface are calculated. The conservation of calculation format is guaranteed. The case study shows that under the condition of rapid change of cross-section geometry of natural rivers， the calculation results of the proposed method are in good agreement with the measured values， and it has high simulation accuracy for mixed flow. The research results provide a new method for the high precision numerical simulation of natural river hydrodynamics and environmental hydraulics.

The traditional water environment capacity is calculated based on 90% of the designed low flow， having the disadvantages such as single hydrological calculation conditions and hard to control according to the water environment capacity results. This paper considers the monthly changes of water environment capacity under the influence of river flow and pollutants entering the river， and proposes a calculation method for dynamic water environment capacity and river pollution control based on changes in hydrological processes and pollution sources. Taking the Jinzhong section of the Xiaohe River Basin in Shanxi province as an example， the monthly water environmental capacity and the remaining amount of the two-water functional zone in the study basin are calculated， and the MIKE11 model is used to calculate the monthly allowable dynamic pollution load into the river. The results show that： ① In 2019， the COD and NH₃-N water environmental capacity thresholds of Water Function Zone A are 138.37～216.54 t and 6.38～14.19 t， respectively， and the water environment capacity is relatively large in July to October； the COD and NH₃-N water environmental capacity thresholds of Water Function Zone B are respectively 417.67～821.05 t and 9.32～29.49 t， and its water environment capacity is relatively small from May to August. ② On the basis of meeting the water quality target of the water function zone， the allowable river inflow load of Water Function Zone A in 2019 is greater than the current river inflow， with more margin of COD and NH₃-N in the flood season； the NH₃-N should be reduced by 0.003 and 0.086 t， should be reduced by 0.05 and 0.143 t in Water Function Zone B in January and February respectively. The allowable pollution load from April to July is greater than the actual river discharge， which can be appropriately relaxed. The dynamic water environment capacity calculated in this paper can objectively reflect the actual water environment capacity and the remaining amount under changing hydrological conditions， and gives the allowable threshold of pollutants entering the river under the premise of meeting the water quality target of the river， providing an important reference and new ideas for the water environment protection and management of the river basin.

Water environment risk assessment is the basis of water environment pollution prevention and risk management. Based on the current situation of the main rivers into Baiyangdian Lake in the upper Baiyangdian Basin， considering the risk sources， river characteristics， physical geography and social development conditions， water pollution control ability and other factors， the water environment risk assessment system of the rivers into Baiyangdian Lake in the upper Baiyangdian Basin is constructed， the water environment risk of the rivers in the upper Baiyangdian Basin is evaluated by using the improved analytic hierarchy process （AHP） and the weighted summation method of comprehensive index. The results show that： in 2017， the overall risk level of the rivers into Baiyangdian Lake in the upper Baiyangdian Basin are in medium risk level， and some sections of Xiaoyi River are at the high risk level； evaluation indicators such as the proportion of environmental protection investment in GDP， the amount of chemical fertilizer applied per unit of arable land， and the current water quality of the river are the main factors that affect the water environmental risk level of the rivers into Baiyangdian Lake in the upper Baiyangdian Basin.The main measures to improve the water environment of the rivers in the upper Baiyangdian Basin include increasing the proportion of environmental protection investment in GDP， properly dealing with the relationship between economic growth and agricultural non-point source pollution， reducing the application intensity of chemical fertilizer and strengthening river governance. The research can provide reference for the treatment and restoration of the ecological environment of the rivers into Baiyangdian Lake in the upper Baiyangdian Basin.

To explore four quite underwater plants in the high concentration of nitrogen and phosphorus stress tolerance ability of the polluted water， artificial simulated water conditions on the dynamic change of nitrogen and phosphorus，nitrogen and phosphorus in the outdoor tolerance pot experiment， through the study of plant antioxidant enzyme defense system main enzymology indexes superoxide dismutase （SOD） and catalase （CAT） activity and malondialdehyde （MDA） content， this paper determines the tolerance range of nitrogen and phosphorus of four emergent plants. The results show that with the increase in nitrogen and phosphorus concentration， the activities of SOD and CAT in the four plants increased first and then decreased， while the content of MDA gradually increased and began to decline after exceeding the plant tolerance capacity. The final conclusion is that the tolerance range of the plants to ammonia nitrogen is as follows： calamus and cattail： 20~400 mg/L， canna and reed： 20~500 mg/L， phosphorus tolerance range was as follows： calamus： 2~200 mg/L， canna， reed and cattail： 2~300 mg/L.

In the downstream of the dam， the tail of the stilling pool or the slope of hydraulic structures， the foundation scouring damage often occurs， threatening the safety of the building， so the anti-impact pile is often used to protect the foundation in engineering. In order to explore the anti-impact characteristics of a new type of anti-impact pile， namely non-bearing pile， this paper uses model test method to measure the actual sediment starting velocity of pileless and non-bearing pile at different lengths from 0.06m to 0.13m by means of propeller current meters， and compares it with the theoretical starting velocity. Meanwhile， the comprehensive stability coefficient of the pile is calculated. The results show that the longer the pile length is， the greater the actual and theoretical starting velocity of sediment is， and the greater the comprehensive stability coefficient is. In addition， the comprehensive stability coefficient of the pile increases with the increase in water depth. The research results of this paper can provide reference for alleviating foundation scouring and protecting flood discharge safety of river dam.

Land use/cover change is the most direct expression of the influence of human activities on the ecosystem. The change of land use will cause the fragmentation， degradation and even disappearance of the habitat patches， which will directly affect the quality of the habitat in a region. Based on an analysis of the land use change from 1986 to 2016 in Qilian mountain areas in the upper reaches of Shiyang River， this paper evaluates its habitat quality. The results show that the main land use types in the upper reaches of the Shiyang River are grassland， woodland and unused land， the main habitat degradation types are non-obvious degradation and slight degradation， and most of the areas are medium or above habitat types， mainly in the west and south. From 1986 to 2000， the area of degraded habitat and the area of poor habitat quality in the upper reaches of Shiyang River increases， but the rate of increase gradually slows down， and the area of good habitat types increased after 2000. Therefore， the habitat quality in the upper reaches of the Shiyang River has a tendency to develop in a good direction.

Suspended sediment is an important carrier of river biogenic material transportation. Clarifying the temporal change characteristics of suspended sediment transport rates in rivers can provide basic information for understanding river ecological rhythms. This paper studies the monthly monitoring data of suspended sediment transport rates and discharge from 1956 to 2000 at the Nugesha， Yangcun， and Nuxia hydrological monitoring stations in the middle reaches of the Yarlung Zangbo River. The missing periods are interpolated by using an established water-sediment relationship. Morlet wavelet transforms， concentration degree and concentration period analysis methods are then used to explore the characteristics of sediment transport changes under different time scales in the watershed. The results show that the suspended sediment transport rates series from the three stations all assumed periodic changes in multiple time scales， maintaining a highly spatial consistency with their four major cycles being 19 years， 26 years， 10 years， and 6 years， respectively. There is a certain response between the short-period changes of sediment transport rates and the periodic changes of solar activity and climate events. The maximum sediment transport rate of the three stations mainly occurs in July and August， and the concentration degrees are 0.83， 0.77， 0.74， respectively. Affected by differences of local meteorological condition， the uneven distribution of sediment transport rates during the year gradually decrease along the river. Moreover， the sediment transport rate showed an abrupt change in 1964 when an increasing trend changed into a decreasing trend. Precipitation and human activity are believed to contribute to this reduction， with the contribution rate from precipitation in the three stations were 38.62%， 41.04%， and -38.15%， respectively， and the contribution rate from human activity were 61.38%， 58.96%， and 138.15%， respectively. This study preliminarily clarifies the multi-time-scale variation characteristics and main control factors of the suspended sediments transport process in the middle reaches of the Yarlung Zanbo River when there are no obvious disturbs from large-scale water conservancy projects， which can provide a natural state reference system for ecological protection in the Yarlung Zangbo River.

The calculation of storage and flood control capacity of river network is of great significance to the scientific storage and rational regulation of river water resources. This paper takes the regional river network of Huantai County as the research object， and constructs a structure decomposition model of river water resources. Based on the natural attributes of different water quantities， the method divides the whole volume of river water into four parts： namely environment water portion （EWP）， safety water portion （SWP）， risk water portion （RWP） and disaster water portion （DWP） with the characteristic water level of ecological water level， waterlogging level， flood control design water level and dike top elevation. In order to maximize the natural “storage tank” function of river network and explore the storage and flood control capacity of regional river network under different scenarios， combined with the river water division model of Huantai County， this paper selects four river network storage and flood control capacity indexes （storage and flood control capacity， regulation storage capacity， storage and flood control capacity per unit area and regulation storage capacity per unit area） to calculate the storage and flood control capacity under ecological operation， resource operation and flood control operation of river network in Huantai County. The result shows that the calculation method of storage and flood control capacity of river network based on river water division can fully reflect the interval attribute of storage and flood control capacity， and lay a good foundation for the selection and implementation of different regulation schemes in Huantai County.

To improve the prediction accuracy of runoff time series this paper proposes a runoff time series prediction method that combines wavelet packet decomposition （WPD） with singular spectrum decomposition （SSA）-rat swarm optimization （RSO） algorithm-echo state network （ESN）. WPD and SSA is used to decompose the non-stationary runoff time series into several sub-sequences to effectively reduce the complexity of the runoff time series； the principle of the RSO algorithm is introduced， 6 typical functions are selected under different dimensional conditions to simulate the RSO algorithm； RSO algorithm is used to optimize hyperparameters such as ESN reserve pool size and sparsity， WPD-RSO-ESN， SSA-RSO-ESN models are established， and WPD-RSO-SVM， WPD-ESN， WPD-SVM and SSA-RSO- SVM， SSA-ESN are constructed， and SSA-SVM are used as comparative analysis models； the monthly runoff time series data from 1957 to 2014 at Jiangbian Street Hydrological Station in Yunnan Province are used to test and compare 8 models. The results show that the RSO algorithm has better optimization accuracy and global search ability under different dimensional conditions. The WPD-RSO-ESN and SSA-RSO-ESN models have predicted average absolute percentage errors of monthly runoff time series for 10 years and 120 months after the example. The average absolute percentage errors are 2.73% and 3.90%， respectively. The prediction accuracy is better than other models under the same decomposition conditions. The RSO algorithm can effectively optimize the hyperparameters of the ESN network and improve the prediction performance of the ESN network. The decomposition effect of WPD on runoff time series data is better than SSA method.

In order to scientifically and objectively evaluate the water resources security situation and on the basis of fully analyzing the actual situation of water resources characteristics in the center of Baoding City，an evaluation index system of water resources security based on the concept of pressure-state-response （PSR） model combined with the methods of correlation analysis and principal component analysis is intended to build. At the same time， an evaluation model of water resources security is established based on the adaptive neuro-fuzzy inference system （ANFIS）. The water resources security situation of Baoding City from 2009 to 2018 is evaluated. The results show that the water resources security situation of Baoding City had no obvious change trend from 2009 to 2018. Except for the main urban area， the water resources security in other areas is mainly in the critical state， and the situation is not optimistic. The state security of water resources is the most important factor affecting the water resources security in the center of Baoding City.

In order to effectively improve the prediction accuracy of non-stationary and non-linear runoff series， the TVF-EMD-LSTM prediction model is proposed by coupling the time varying filter empirical mode decomposition （TVF-EMD） with adaptive sequence characteristics and Long Short-Term Memory neural network （LSTM）. Firstly， the TVF-EMD method is used to adaptively decompose the runoff series into high-frequency series and low-frequency series； secondly， the LSTM neural network is used to predict the decomposed series. Finally， the predicted results are added and reconstructed into the final runoff prediction results. The proposed model is applied to the monthly runoff forecast of the Changshui Hydrological Station in the Luohe River Basin， and compared with LSTM model， EMD-LSTM model and CEEMDAN-LSTM model. The results show that the TVF-EMD-LSTM neural network coupling model has the highest prediction accuracy and the smallest error. Therefore， TVF-EMD can better alleviate the mode mixing problem and provide a better way for the data preprocessing of runoff series. The TVF-EMD-LSTM hybrid model provides an effective new method for monthly runoff forecasting.

In order to divide runoff of urban river without upstream incoming water and further explore the impact of urban underlying surface on rainfall-runoff relationship， several automatic division methods of flow process line are introduced to divide the daily runoff of Zhongmu Hydrological Station from 2003 to 2016， then the urban river runoff is divided into annual base flow and annual direct runoff by using the selected optimal division method. The annual direct runoff coefficient is used to express the urban water catchment characteristics， and the five-phase land use data are combined to construct a multiple linear regression model of the annual direct runoff coefficient and various underlying surface areas， the annual direct runoff coefficients for 2017 and 2018 are predicted. The results show that the digital filtering method F1 has the best division effect on urban river runoff without upstream incoming water； the contribution rates of base flow and direct runoff to the total runoff growth are 83.96% and 16.04% respectively； the multiple linear regression model is significant at 0.01 level and has good performance， where NSE， R ²， and R_e are 0.891， 0.891 and 6.72% respectively. Moreover， the relative errors of annual direct runoff coefficient forecast for 2017 and 2018 are 4.80% and 9.75% respectively.

H-ADCP is widely used in China. Especially in the middle and lower reaches of the Yangtze River， it has become an important part of hydrological flow measurement. However， there is a general lack of analysis and verification of the original data of flow measurement. That leads to low accuracy and reliability. And it also makes comparison test and calibration more complicated. In view of the above problems， and according to the principle of H-ADCP， combined with the hydrological characteristics of plain river network area and tidal reach， this paper designs a high credibility H-ADCP flow automatic monitoring and analysis processing system based on domestic measurement and control platform. This system can analyze and verify the original data of the flow measurement， automatic lifting of flow measurement bracket， comparison test and calibration. By application， this system can effectively identify and eliminate anomalous flow measurement data， improving the accuracy of flow monitoring results. It’s worth popularizing and applying. At last， this paper prospects two types of new technology： rotary scanning and multi-level joint scanning.

An investigation of the effects of annual runoff wet-dry classification methods is of important significance for hydrological analysis and water resources management in a river basin. Three corresponding indexes， i.e. similarity index， difference index and uniformity index， are defined for evaluating the wet-dry classification effects from the view of annual runoff process similarity， volume difference and distribution uniformity of wet-dry categories. According to the adopted classification index types， annual runoff wet-dry classification methods are divided into three categories， i.e. comprehensive index-based （CI-based） category， relationship index-based （RI-based） category and comprehensive relationship index-based （CRI-based） category. Moreover， nine representative methods are selected from the three categories， and each category contains three methods. Finally， Yingluoxia Station of Heihe River Basin is taken as a case study， where the selected nine methods are applied to the annual runoff wet-dry classification of the station， and the wet-dry classification effects of different methods are analyzed. The results show that ① annual runoff wet-dry classification results are very sensitive to classification methods and their internal parameters， classification results of different methods are obviously different， and the classification effects of the same method may also differ from region to region； ② the three defined indexes may reflect the wet-dry classification effects of different methods， and the uniform distribution of wet-dry classes is beneficial for the improvement of wet-dry classification effects； ③ overall， the classification effects of the methods of CRI-based category are the best， those of CI-based category are the second， and those of RI-based category are the worst； and ④ the classification results of conventional methods may be more reasonable than those of comprehensive methods when the annual runoff volume and process are relatively consistent. The results can provide a scientific basis for reasonably classifying annual runoff wet-dry states of a river basin.

Traditional annual runoff classification is often divided into wet years， normal years and dry years by formulating quantitative indicators of annual runoff. However， it is difficult to reflect the situation of annual runoff only by dividing the results of annual runoff. This paper introduces the SOM neural network， and divides the annual runoff sequence according to the monthly runoff of the runoff sequence， the average monthly runoff during the year， the uneven coefficient of the year and other indicators， and obtains the results of the annual runoff division with the characteristics of the year. Taking Miyun Reservoir as an example， the annual runoff of Mi-Yun Reservoir is applied to research， and the annual runoff sequence of Miyun Reservoir is divided into 3 categories： wet and normal and dry， and then the results are subdivided into 9 categories. The results indicate that different classification results have different properties in runoff， intra-year distribution， unevenness， concentration， and intra-year change range. The representative years of the 9 sub-categories are similar in runoff in the same major category， and in the runoff distribution， the runoff in the dry season has its own characteristics and different distributions. It shows that the SOM neural network considering multiple indicators can divide the annual runoff sequences in detail.

This study combines hydrological statistical method， daily drought calculation method and its theory to quantify the drought and flooding events through historical observed meteorological data. These events are used to analyze the historical climate evolution and drought and flood in Wuhan. The results show that the historical rainfall in Wuhan increases and evaporation decreases. The evaporation is greater than rainfall in the dry season， which indicates that the water level of the lake might decrease in the dry season. In addition， the abrupt change of drought/flood in summer and the abrupt change of flood/drought in autumn and winter may lead to the risk of lake water level fluctuation. This paper provides not only a reference for the changes of rainfall evaporation and the water level control plan for the lakes in Wuhan， but also provides the corresponding technical basis for the overall prevention plan of extreme drought and flood and its abrupt change in Wuhan and even the Yangtze River Basin.

In view of the dynamic variation of unit efficiency coefficient and its value in the middle and long-term power generation dispatching of the reservoir supported by the tailwater downstream， this paper starts with the reservoir water level， tailwater level and power generation flow of the reservoir， the key factors affecting the comprehensive efficiency coefficient under different water conditions and their influences are analyzed. A case study is carried out combined with the hourly operation process of the Three Gorges Hydropower Station. The results show that the ten-day comprehensive efficiency coefficient has a dynamic change， the difference between the maximum and the minimum is 9.77% （8.71% in flood season）， and there is also a certain inter-annual difference （the minimum in 2017 is 4.44% more than that in 2018）. A further analysis shows that the value of ten-day comprehensive efficiency coefficient is smaller in the period of abundant water， the power generation flow is the main influencing factor， the tail water level takes the second place， the reservoir water level ranks lastly， the value is larger in the period of moderate or small discharge， and the power generation flow is the main influencing factor， the reservoir water level takes the second place and the tail water level ranks lastly.

Taking a ground surface steel penstock of a hydropower station as an example， the finite element model of the whole hydropower station is established. The effects of uniform temperature and non-uniform temperature on the stress and deformation of the structure are considered， and the feasibility of canceling the expansion joint is demonstrated. The results show that the Mises stress of steel penstock after canceling the expansion joint is less than the allowable stress value of the steel. The cancellation of the expansion joint makes the steel penstock in the upper horizontal section produce large axial stress， but it does not cause the buckling instability of the steel penstock. From the aspects of structural strength safety and stability， it shows that it is possible to cancel the expansion joint of the ground surface steel penstock. The mechanism of the penstock under different temperature after canceling the expansion joint is further explored. The results show that the non-uniform temperature has a great influence on the axial stress of the penstock， which may cause the axial buckling stability of the penstock. Therefore， it is necessary to pay attention to the axial buckling stability of the steel pipe in addition to the strength check.

The safety of high-core rockfill dams has sparked widespread concern due to their complex environments， large loads and high stress levels. However， crest cracking occurs frequently in practical engineering， which affects the anti-seepage and stability of dams， and it is of great significance to analyze and study the relevant factors affecting the dam crest cracking for the design and operation of dams. Based on this， by comparing the variation law of deformation gradient of dam crests before and after the cracking of typical dams， the deformation gradient is introduced as an index to analyze the influencing factors of dam crest cracking of high-core rockfill dams. Based on the relevant information of several built high-core rockfill dams， a research scheme considering the size of the dam body， the thickness of the overburden， the height and rate of water storage， and operation time is constructed. The results show that the deformation gradient of dam crest increases significantly with the increase in dam height， and the relationship between them is nonlinear. The dam crest cracking risk of high-core rockfill dams built on deep overburden is greater， but when the overburden thickness reaches more than 100 m， the deformation gradient of dam crests tends to converge. The deformation gradient of dam crests is more sensitive to high water storage level， and the deformation gradient of dam crest increases with the acceleration of water storage rate.

The fine parametric modeling of equipment is driven by database based on French Dassault 3DE collaborative design platform by relying on the design work of hydraulic machinery specialty of Wudongde Hydropower Station. After electromechanical interface is defined， the mechanical and electrical equipment model library is built. Through the interactive and collaborative design of hydraulic machinery， electrical， HVAC， civil engineering and so on， multi-specialty interference check and establishment of automatic avoidance rules， the construction of 3D model is completed. The 3D design is driven through the system schematic diagram. After 2D drawing is created， it can mark parts and generate material list automatically. Then EKL language of 3DE platform is used to customize material report and other BIM design technologies， this paper combs out a set of hydraulic machinery full forward BIM design process.

Lining of hydraulic tunnel is the main component of tunnels， and it works together with surrounding rocks to support loading， and then enhances the stability of surrounding rock. How to model the performance of the lining of the hydraulic tunnel is worth discussing. By taking a hydraulic tunnel in a water diversion project as an example， a hybrid model of steel bar stress meter is put forward for the lining of hydraulic tunnels based on in-situ data. The model is composed of internal water pressure， temperature and timing component. And internal water pressure is calculated by structure formula， temperature and timing are acquired by statistics. Therefore， measured stress is well reflected by fitting curves， correlation coefficient of R6 and R5 between fitting value and measured value is 0.987 and 0.861 for inner and outer steel bar stress respectively. Furthermore， monitoring warning value for inner and outer steel bar stress is preliminarily established. Above all， hybrid model put forward herein can be used for monitoring the change of lining of hydraulic tunnel， and provide reference for the operation of projects.

The formation mechanism and mechanical characteristics of landslides are an important basis for guiding landslide prevention strategies. In the current landslide simulation， the hypothesis of soil strength weakening cannot reflect the structural damage characteristics and the weakening of deformation parameters are not considered. Thus mechanical characteristics revealed are distorted. This paper suggests using stress level instead of plastic shear strain to characterize the degree of structural damage to construct a new strength reduction method. Based on this， the progressive failure process， stress and displacement characteristics and stability attenuation laws of different types of landslides are analyzed， which reveals the cause mechanism of different types of landslides. Main conclusions are as follows. ① The main sliding section of the push type landslide is in the middle of the slope， the pull section is in the upper part of the slope， and the anti-sliding section is in the lower part of the slope. The main sliding section is destroyed first. Then failure gradually expands to the upper and lower parts. Finally a shear failure surface is formed， on which the shear strength is fully exerted. ② The main sliding section of the pull-type landslide is at the foot of the slope， the pull section is in the middle of the slope， and the anti-sliding section is in the upper part of the slope. After the shear sliding of the main sliding section， the tension-shear sliding occurs quickly at the pull-type section and anti-sliding section because of losing the support from the main sliding section. And the shear strength of the rock and soil mass in the upper part of the pull-type landslide is not fully exerted. ③ For the pull-type landslide and push landslide whose initial stability state is similar， the stability state of the pull-type landslide decreases rapidly after the occurrence of large shear deformation， and failure of pull-type landslides happens suddenly and uncontrollably. But the push landslide is under metastable state for a long time before failure. So there is a long buffer time for push type landslides to reinforcement treatment. ④ From the perspective of disaster prevention， the push type landslide should pay attention to “strengthening the waist”， and the pull-type landslide should pay attention to “strengthening the foot”.

Seepage flow monitoring is an important project of dam safety monitoring. At present， manual measurement method is mainly used to monitor the seepage flow less than 1 L/s， which is time-consuming and laborious， and cannot realize remote online real-time monitoring. Therefore， a small flow automatic online monitoring seepage flow meter based on volumetric measurement principle is designed. Through the design of the measuring circuit， the influence of the structure of the instrument on the stability of the measuring device is avoided. Manual comparative measurement proves that the instrument has the characteristics of high measurement accuracy and long-term stability. It is suitable for the automatic on-line monitoring of dam seepage in small flow， multi-impurity and humid environment. It saves manpower and material resources and has good application value.

The dynamic response and failure mode of earth rockfill dam under earthquake is always a research hotspot in geotechnical engineering. FLAC finite difference software is used to analyze the residual deformation， acceleration response and shear strain of Shanmei earth rock dam after earthquake， and the general dynamic response law of Shanmei earth rock dam is obtained. By changing the upstream water level， the evolution mode of plastic shear strain area in the dam during the earthquake and the residual strain pattern in the dam after the earthquake at different water levels are described. It is found that the shear strain concentration position of Shanmei Earth Rock Dam is generally at the junction of different dam materials at the dam crest and dam slope. The shear strain pattern in the dam is basically unchanged after a stationary earthquake， but the shear strain is continuously accumulated， In addition， the gradual elevation of water level will lead to the formation of a plastic discontinuous shear zone in the upstream dam material after the earthquake. This study helps to understand the cause and formation process of plastic shear strain zone of Shanmei Earth Rock Dam under earthquake action， and can provide an important reference for the seismic work of Shanmei Earth Rock Dam.

In order to solve the difficulties in controlling the turning angular velocity leading to the angle measurement ferrying data with the corresponding problem by using eddy current sensor and new type of grating sensor device， using an adaptive variable speed barring consecutive ferry method is used， and the development of analysis and processing system is based on Matlab， the accurate calculation of the pump shaft fixed point on the ferry， and the corresponding axis adjustment scheme is given for your reference. Relevant achievements can be applied to turning and swing measurement of large vertical pump stations.

Based on chaos theory and nonlinear dynamics， the fault warning diagnosis method is proposed where the maximum Lyapunov exponent is the evaluation index for the operating state， and the Elman-decision tree is used to predict and classify the failures. The method is testified by using a data set proposed by Paderborn University， which is measured from a normal bearing， an outer ring artificially Electrical Discharge Machined （EDM） bearing and outer ring drilled bearing. The result shows that the maximum Lyapunov exponent can effectively distinguish the data sets from different fault states. Then an Elman-decision tree model is established and trained to realize the maximum Lyapunov exponent prediction and fault identification， which is proved of a relatively high accuracy based on the simulation results. The proposed method can achieve good performance even in presence of moderate noise， and can reduce the computation burden by avoiding training multiple corresponding neural networks for all the faults， and it provides a new idea for fault warning diagnosis.

In China， the inadequate operation management system and low informatization have led to the economic benefit attenuation and potential security risks of small reservoirs. On the basis of sorting out the current situation and problems in small reservoir management， the characteristics and focus of safe operation and management of small reservoirs are summarized. Four principles that should be followed in safety operation and management of small reservoir are clarified， several suggestions are made for its management. Based on the management mode of small reservoirs in Luquan District， Shijiazhuang， it is proposed that the property-based and information-based management mode can enhance the management efficiency of small reservoirs， which serves as a reference for the realization of safe and long-term operation of small reservoirs.

Identifying important nodes in water delivery system plays a significant role in its protection resources distribution and safe operation. A topology network of long-distance water delivery system is constructed by using complex network theory， then a node importance evaluation is achieved with the consideration of local centrality， closeness centrality， betweenness centrality and network centrality from local， global and location aspects. Then evolution process of water delivery network comprehensive efficiency is calculated by removing nodes following their degree of importance. The results show that the nodes located in the middle of the water delivery system are more important， the water delivery network comprehensive efficiency decreased from 0.12 to 0.02 while 34% important nodes were destroyed， which indicates that a few important nodes failed can cause the rapidly collapse of the water delivery system.

Accurate prediction of water supply is one of the key factors for the efficient and stable operation of urban water supply systems. In particular， the accurate forecast of short-term precipitation will help narrow the gap between production and sales and significantly reduce energy consumption. This paper proposes an artificial bee colony optimization multi-variable gated recurrent unit model （WD-ABC-MGRU） based on wavelet decomposition， which can effectively capture multi-variable and multi-scale time series characteristics， thereby significantly improving the model’s ability to handle complex patterns and achieve accurate water supply prediction. WD-ABC-MGRU selects the strongly correlated factors affecting water supply， combines the components of water supply obtained by wavelet decomposition and coefficient reconstruction and performs phase space reconstruction to form the model input， builds the model separately for prediction， and sums the predicted values of each model to obtain the final water supply prediction result. The experimental results show that the WD-ABC-MGRU model provides more accurate water supply forecasts than the GRU， MGRU， WD-MGRU and WD-ABC-BP models， with an R ² of 0.87 and a MAPE of 1.75%. It can be seen that this model can provide a practical solution for the accurate prediction of urban water supply， thereby contributing to the efficient and stable operation of the urban water supply system.

Considering the necessity of continental water diversion in island areas and the complexity of water supply process through water engineering groups， this paper proposes a method of optimal allocation of water resources suitable for island areas. Based on the optimal allocation concept of “zoning and grading”， this paper analyzes the hydraulic connection under the complex system of the island area and carries out the system generalization to complete the “zoning”， and then combined with the characteristics of the water plant to complete the “grading”. According to the multiple water sources and multiple users， this paper develops a multi-objective optimal allocation model with maximizing water supply guarantee rate， minimizing water supply cost and minimizing water plant residual storage capacity as objective functions and then uses the multi-attribute decision making method which combines subjective and objective weights to deal with the Pareto solution sets helping to get solutions scientifically by taking the water resource allocation of Zhoushan island as an example. Results show that the multi-objective optimal allocation model of complex water engineering group in island area based on the concept of “zoning and grading” can improve the water utilization efficiency of the island and reduce the water supply cost by 4.5 million yuan per year. The optimized model can dynamically adjust the island water and mainland water supply ratio under different water inflow conditions. Multi-attribute decision-making can assist decision-makers to optimize the scheme under different scenarios， provide a scientific basis for the optimal allocation of water resources in island areas， and improve the comprehensive benefits of society and economy.

It is of great significance to reduce the underground water level at the bottom and both sides of the channel to ensure the stability of the slope and the safety of the large open channel. Taking the Laizhou section of the Yellow River Water Diversion Project in Jiaodong， Shandong Province as the research object， a two-dimensional hydro-geological model and a seepage model of the channel slope are established to quantitatively analyze the pressure relief effect of drainage pipes with different heights and horizontal distances on both sides of the channel. According to the variation law of total water head and pore water pressure at different drainage pipe positions， the optimal drainage and pressure reduction position is determined. The results show that the best location for the effect of drainage is 0.1~0.125 m from the side of the canal. In addition， the drainage pipe positions of different sedimentation carrying capacity channels are studied， which serves as a basis for the reasonable drainage pipe arrangement in different channels， maximizing drainage and depressurization effect and preventing uplift pressure damage of channels.

Fiber Bragg grating （FBG） sensors have the characteristics of anti-electromagnetic interference， no power supply， corrosion resistance and high sensitivity. As a common construction structure in water conservancy industry， diversion tunnel has some hidden safety problems due to its structure， engineering and environment. Therefore， it is necessary to carry out long-term effective stability monitoring. Based on the water supply engineering of 7# tunnel in Gansu province as an example， the optical fiber sensing technology was introduced in the safety monitoring of the diversion tunnel structure. FBG sensors were used in different encapsulation structure to monitor the stress， the displacement， the temperature， and the settlement of the tunnel effectively. The measured data also verified the reliability of the application of optical fiber sensing technology in tunnels.