To ensure the safe and stable operation of large radial gate in service, taking the hydraulic radial gate of Gezhouba Project as an example, the finite element analysis software was used to analyze the static strength and modal to obtain the maximum stress concentration point and deformation of the gate under the design head condition. The vibration intensity of the gate was judged according to the natural frequency, vibration model and maximum deformation displacement. The results show that the deviation between the total design water pressure of the gate and the theoretical calculation is 4.7%, which is located in the 5% error control range. The overall maximum stress of the radial gate is 373.86 MPa and the deformation is 12.938 mm, meeting the design requirements of the specification. The bottom of the panel exists local stress concentration, the overall vibration deformation of the gate is less than 0.508 mm, and the vibration intensity is small. Only low-order frequencies appear in the high-energy region of the pulsating pressure of the water flow, and the possibility of resonance is small.

The space angles of inclined supporting arm of large radial gate are easily confused and mistaken, and its small errors may cause large structural errors, which will affect the manufacturing, installation and safe operation of the radial gate. Therefore, this paper defined, distinguished and calculated several key space angles of the radial gate inclined supporting arm. Taking an actual project radial gate as an example, the influence extent of each key angle of the inclined supporting arm on the manufacturing and installation accuracy of the gate was analyzed. Furthermore, this paper put forward several measures to improve the manufacturing and installation accuracy, which can be used as a reference for the subsequent design, manufacturing and installation of radial gate with inclined supporting arm.

As the most basic precipitation collection equipment in the field of hydrometeorology, the tipping bucket rain gauge has the characteristics of practicality and good stability, but there are some deficiencies in the research on the tipping characteristics and mechanism of the tipping bucket rain gauge and the causes of the mechanical error of the tipping bucket rain gauge. In this study, the mechanism of the tipping process of tipping bucket rain gauge was explored by means of combination of numerical simulation and indoor experiment with high-speed camera. The results show that during the turnover process, the time from start to stop is about 0.35-0.40 s, the time t₁ from start to level is relatively high, about 80%-85%, and the time t₂ from level to stop is only about 0.05 s, and there is a slight rebound after the turnover touches the adjusting nut. During t₁ period, water will be continuously injected into the tipping bucket, which is the fundamental reason for the change of additional water volume and mechanical error of rain gauge with rain intensity. The tipping bucket will pour out the water in a very short time (about 0.4 s), and because the tipping bucket will cause slight rebound when touching the stop bolt, it will cause the "double peak" phenomenon of the tipping flow of the rain gauge tipping bucket when pouring out the water, and it will increase with the small installation inclination of the tipping bucket on the left and right. The research results can provide basic theoretical support for the improvement of tipping bucket rain gauge in the future.

In the hydraulic transition calculation of long-distance water transmission pipeline, the intake and exhaust coefficient of air valve is mostly constant. In order to explore the influence of the intake and exhaust coefficient of water hammer air valve on the change of pipeline pressure, the DN100 water hammer air valve was simulated by FLUENT software to obtain the mass flow at -9 kPa -11 kPa, and the dynamic intake and exhaust coefficient curves with different pressure differences were established. Taking Xishan primary water lift pump station as an example, dynamic value and fixed value were adopted for air valve inlet and exhaust coefficient respectively (mean value 0.55) to calculate and compare the transition process. The numerical simulation results show that the improved dynamic inlet and exhaust coefficient of the air valve has little impact on the negative pressure of the pipeline and great impact on the positive pressure. The improved dynamic inlet and exhaust coefficient effectively reduces the pressure fluctuation in the pipeline. In the actual operation of the pump station, it is necessary to pay attention to the positive pressure change of the pipeline, and strengthen protection. The calculation of water hammer prevention should be based on the measured data of air valve inlet and exhaust. The research results can provide some references for the hydraulic transition calculation of air valve in the operation of pump station.

Aiming at the inability to accurately locate the fault of the transmission line in hydropower station, an improved double-end location algorithm independent of wave velocity was derived from the actual propagation theory of traveling waves. Using the unique advantages of variational mode decomposition (VMD) which can adaptively decompose fault current traveling wave signals, and the Teager energy operator (TEO) which can quickly track the energy change of signals and has less computation, a traveling wave positioning model for 220 kV transmission lines based on VMD-TEO was established. The model does not need to consider wave velocity correction and double terminal time synchronization. The simulation results of PSCAD show that the maximum error percentage of the improved double-end fault location model is 0.239%, which is superior to other location algorithms. The reliability and superiority of the proposed method were verified.

In practical engineering, there are few samples of arrester failures, and it is difficult for intelligent algorithms such as neural networks to make accurate judgments. To this end, a fault diagnosis method of arrester based on Bayesian network was proposed. Firstly, the principal component analysis was used to extract 21 characteristic parameters that affect the operation of arrester. And then the extracted characteristic parameters was chosen to establish two-layer information architecture defect diagnosis model. The classification probability of different categories was calculated according to the existing real-time data. If the first classification result indicated that the arrester is abnormal, new detection evidence was added for the second diagnosis. Finally, 6 arresters under the same voltage level in a certain area were selected to analyze and verify the validity and correctness of the proposed method.

A CVMD-GRU-DenseNet model for short-term load forecasting based on a decomposition-prediction-reconstruction framework is proposed to aim at the nonlinear and multi-period characteristics of power load time series. In the decomposition stage, the optimal decomposition number of VMD is determined according to the correlation entropy between subsequences to improve the decomposition quality. In the prediction stage, the input features are selected according to the characteristics of each sub-sequence. The GRU neural network and the DenseNet model are employed to forecast the regular low-frequency and highly random high-frequency components, respectively. Finally, the prediction results for each element are reconstructed into a load prediction curve. The short-term load forecasting results of four seasons for a city in Hubei Province show that the proposed method can effectively improve forecasting accuracy and has strong generalization ability.

Aiming at the topology optimization of large-scale offshore wind power collection system, an improved partheno genetic algorithm was proposed. The calculation model of the present value of the life cycle cost of the power collection system was established. Taking the minimum of the sum of present value as the goal, the population generation mode meeting the current carrying capacity constraints of the submarine cable and the elimination coefficient related to the number of crossing points were designed to avoid the submarine cable crossing. The topology of the power collection system with 35 kV and 66 kV voltage levels was optimized and compared. The example results show that the total cost of the improved partheno genetic algorithm can be reduced by 14.2% compared with the traditional algorithm; Compared with 35 kV, 66 kV voltage level has better comprehensive benefits, which can provide reference for the planning and design of large-scale offshore wind farm power collection system.

In order to improve the prediction accuracy of ultra short-term wind power, a combined wind power prediction model based on IEWT-FE-BO-LSTM was proposed. Firstly, an improved empirical wavelet decomposition (IEWT) was used to decompose the historical wind power data. The Fuzzy Entropy (FE) algorithm was introduced to calculate the complexity of each decomposed submodel and reconstruct the submodel. For each recostructed component, a prediction model based on long short term neural network (LSTM) was established. Bayesian optimization algorithm (BO) was used for hyper parameters to solve the problem of poor training results caused by manual parameter adjustment. The example analysis based on historical wind farm data shows that the IEWT-FE-BO-LSTM model has high prediction accuracy and efficiency for ultra short-term wind power.