The article addresses the problem of relatively low accuracy of traditional PV power prediction and proposes a hybrid TOPSISGRNN based mechanismdata driven PV plant power prediction model. Firstly, the correlation analysis of several meteorological indicators and the output power of PV power plant is carried out, and the meteorological data with high correlation is selected as the input factor of the model. The TOPSIS algorithm was used to select the optimal similar days, and then the theoretical values of their PV plant output power and meteorological data were used to build the GRNN prediction model. Finally, the model was simulated and validated by combining the historical meteorological data and power data on the DKASC website. The final test results yielded an average power prediction accuracy of 0.826 9 kW for RMSE, 3.45% for MAPE and 0.019 5 kW for MAE. The prediction accuracy of this forecasting method is significantly higher than that of a single forecasting model and has some theoretical and practical value.

The wind turbines are subjected to continuous wind loads during their whole service life and are inevitably resisted earthquake action. In this paper, a large megawatt wind turbine is taken as the research object to simulate and analyze its deformation and mechanical responses under the conditions of normal operation, earthquake during shutdown and earthquake during operation. The results show that the deformation of wind turbine increases with the increase of tower height, and the horizontal displacement of the wind turbine under the earthquake is obviously less than that under the wind turbine operation state. With the increase of earthquake level, the horizontal displacement of the wind turbine is gradually increased. The deformation shape of the wind turbine remains unchanged, and the deformation of tower top is large. The influence of the wind turbine operating load on structural deformation response is significantly greater than that of small and medium earthquake actions. Meanwhile, the deformation response of the wind turbine under the combined action of operating load and earthquake is significantly greater than that under the separate action of two loads. In the structural design of the wind turbine, it is necessary to consider the coupling effect of the operating state and the earthquake. The stress response at the top of the tower is the largest, and its crosssection should be checked in the design.

In order to make reliable dynamic prediction of the potential risk of pitch system, aiming at the problems of multiple components, complex system and difficult fault feature extraction of pitch system, the fault tree is established through the induction and analysis of its fault point and fault transmission process, and then it is transformed into a dynamic Bayesian network (DBN) integrating Leaky Noisyor nodes, which ensures the accuracy of the model and has the dynamic prediction ability. The model is optimized and verified by using a 5fold crossvalidation method. The test results show that this method has high accuracy in risk prediction, fault cause analysis and risk dynamic evolution process analysis of pitch system, and has engineering application value in guiding the preventive maintenance of pitch system and ensuring the overall safety of wind turbine.

In order to improve the control performance of wind turbine electrohydraulic pitch system, a fractional terminal sliding mode control method based on perturbation observer is proposed. The mathematical model of the wind turbine electrohydraulic pitch system is established, and the slidingmode state and perturbation observer is used to compensate the uncertainty and unknown disturbance of the pitch system parameters in real time. Fractional calculus theory is used to design the sliding mode surface of the terminal sliding mode controller, which can improve the jitter of the sliding mode control itself while ensuring the finite time convergence. Simulink is used for experimental verification, and the results show that the method enhances the antiinterference ability of the pitch system, weakens the jitter of the system, improves the tracking accuracy of the pitch angle, and improves the stability of the pitch system.

Based on the concept of load safety margin, the structure optimization inversion design was complemented for one offshore wind turbine supported by bucket foundation in order to reduce the amount of materials and structural cost while ensuring the strength and stability requirements of the structure. Hence, the optimization feedback analysis model and calculation process of offshore wind turbine supported by bucket foundation has been established and the origin bucket foundation was optimization design considering the three safety margin values of 1.10, 1.20 and 1.30. It can be seen in the results that the optimized foundation structure still has a better safety reserve after considering the safety margin with the design indicators of the bucket foundation structure, which all can meet the design requirements. This research provides a new idea for the optimization design of offshore wind turbine structure.

With the high penetration of renewable energy and electronic equipment, the problem of wideband oscillation in power systems is becoming increasingly prominent, which has become a key factor restricting the efficient consumption of renewable energy. A wideband phasor measurement method is proposed and the corresponding device is developed to satisfy the requirements of wideband oscillation mitigation and protection in power systems with renewable energy. Based on the windowed discrete Fourier transform and three peak interpolation, this device can measure multi –mode wideband phasor of multiple voltages and currents. Besides, the amplitude and phase of high frequency phasor are compensated based on the linear regression, which effectively improves the measurement accuracy of the device while ensuring the dynamic response speed. Finally, the measurement performance of the developed device was verified by experimental testing with testers and realtime digital simulation, and the device can provide support for ensuring the safety and stability of power systems and the consumption of renewable energy.

The integrated energy system (IES) planning and optimization faces multiple challenges such as high volatility of new energy sources and large uncertainty of output. In view of this, this paper proposes a twostage capacitycost planning and optimization method for integrated energy systems considering scenery uncertainty. Firstly, Latin hypercube sampling is applied to generate the base wind and solar scenarios set, and the scenarios are reduced based on the improved kmeans algorithm. Secondly, a multiobjective optimization model is constructed with the lowest operating cost, optimal carbon emission reduction, and optimal pollutant emission reduction; finally, the system capacitycost twostage planning and optimization solution strategy is proposed, and a business park in the south is selected for the planning simulation. The simulation example shows that the twostage planning model of integrated energy system constructed in this paper can ensure the economy of system and environmental protection at the same time, and meet the multiple energy demands of users.

In this paper, a reactive power optimization method for wind farms considering the reactive power resources allocation cost is proposed to achieve the optimal reactive power control under different operating conditions. With the comprehensive cost composed of power loss and reactive power cost as the objective function, the reactive power optimization model is established based on chanceconstrained programming, considering the influence of wind power fluctuations on voltage magnitude, which is then solved by the improved interior point method. According to the actual operating requirements, the proposed method can schedule reactive power of Static Var Generator (SVG), Energy Storage (ES) and wind generators in sequence by setting reactive power cost coefficients, and optimize the power loss of wind farm. The voltage violation risk is also controlled by reserving the voltage safety margin in the voltage constraint. In the end, the advanced and feasibility of the proposed method is verified by a simulation at an actual wind farm in China.

Voltage violation has become an important factor limiting the maximum integration capacity of photovoltaics. To address the issue of voltage violation caused by largescale photovoltaic grid connection, the article proposes a grouped coordinated voltage control strategy for distribution networks with high proportions of photovoltaics. Firstly, based on the different voltage sensitivities of photovoltaic connection nodes in the distribution network, the concept of grouped coordinated control for photovoltaic inverters is introduced. Then, within each group of photovoltaic inverters, voltage control is carried out using capacity utilization ratio and power factor as consistent variables, while intergroup coordination control ensures that the voltage at key nodes converges to the set value of 1.05 p.u.. Finally, through case simulations, the proposed control strategy is verified to effectively suppress voltage violations in distribution networks, avoid unnecessary active power reduction, and demonstrate strong robustness during load and photovoltaic fluctuations.

In order to further improve the efficiency and reliability of the local power grid in the process of fault recovery, this paper proposes an island partition strategy of local power grid with distributed generation based on improved GSAGWO algorithm. Firstly, the optimalworst method is used to evaluate the load to obtain the load weight value, so as to determine the priority of island division of important load restoration under local power grid fault. Secondly, combined with the load priority, the load level weight coefficient is determined, and the objective function model of island division of local power grid with distributed generation is constructed. Then, in order to obtain better objective function solution results, chaotic reverse learning and genetic annealing algorithm (GSA) are introduced to improve the grey wolf optimization algorithm (GWO) to improve the optimization performance of the algorithm. Finally, the modified IEEE 69 node is taken as an example for simulation analysis. The improved GSAGWO algorithm is used to solve the local distribution network fault model, and a better islanding result is obtained. The example analysis shows that the strategy proposed in this paper can accurately realize the optimal strategy of island division under local grid fault, ensure the restoration of power supply of important loads, and verify the effectiveness and superiority of the strategy.