With the continuous increase in the scale of distribution network construction and operation，a large number of electricity load control problems have emerged，especially the weakening of sample time series characteristics in electricity load data analysis，resulting in incomplete acquisition of local load characteristics，making it impossible for the distribution network electricity load prediction system to maintain the balance of energy in the power grid. In the past，the impact of "source-load" uncertainty was also rarely considered in the distribution network electricity load processing and distribution，which can easily lead to low probability of electricity load classification recognition，significant errors in electricity load prediction，and long-term problems such as high generation costs and insufficient distribution balance. In response to the above situation，the uncertainty of "source" and "load" was conducted research through probability distribution functions. A multi-objective function composed of two probability distribution functions was set as the constraint conditions for the coordinated output balance. The improved cluster eddy current search algorithm was used to solve the problem，and a coordinated control scheme for electricity load was obtained. The test results show that at 17：00，the power of photovoltaic device 1 is 1 050 kW， at 11：00，the power of photovoltaic device 2 is 980 kW，based on the source-load uncertainty root-mean-square error of less than 0.7%. The cost of the scheme based on "source-load" uncertainty is 453 200 yuan，and the balance degree is 0.94. The collaborative control method based on "source-load" uncertainty has a higher balance than the collaborative control method without "source-load" uncertainty，and the collaborative control technology is more reasonable.

In response to the potential frequency secondary drop problem of direct-drive wind power systems in the scenario of continuous frequency drop faults in the power grid，the principle of primary frequency regulation of direct-drive wind power systems was analyzed. A primary frequency regulation control strategy for direct-drive wind power systems was proposed based on the product and difference combination of frequency change rates at pre- and post-sampling times. According to the established criteria，the droop coefficient of the direct-drive wind power system in different frequency response stages was adjusted by comprehensively considering the fan speed，frequency deviation，and frequency rate of convergence，to realize the active frequency support and flexible frequency modulation exit of the wind power system. A simulation model of a direct-drive wind power system was established in Matlab/Simulink to verify the effectiveness of the proposed control strategy.

The micro-grid solves the problem that the intermittent and fluctuating power generation has adverse effects on the stable operation of the distribution network when the distributed generation is connected to the distribution network. In order to meet the economic operation of micro-grid under grid connected mode and improve power supply reliability，a micro-grid energy dispatching strategy based on peak-valley price and energy storage state of charge （SOC） was proposed. The strategy divided the whole day into three periods：peak，average and valley. During the real-time scheduling cycle，different scheduling strategies were applied based on different time interval and the SOC of energy storage. Reasonable energy storage charging and discharging penalty functions were designed in different time interval，and the maximum energy storage charging and discharging constraint factor was introduced to further improve the charging and discharging of the energy storage device. The minimum operating cost of micro-grid was took as the objective function and solved it through particle swarm optimization algorithm. The effectiveness of the strategy was verified by an example analysis.

In order to improve the accuracy and stability of short-term wind direction prediction，a wind direction prediction method based on lidar wind data and an improved nonlinear echo state network （NESN） model was proposed.First of all，wind direction data 100 meters ahead of the wind turbine was obtained by laser wind detection radar. Secondly，the multivariate polynomial function was used to construct the nonlinear relation of the internal state of the reserve pool，the order of the weight matrix and the complexity of model calculation were reduced.Finally，the prediction model was established and the simulation prediction was carried out on different lidar data sets.The results show that compared with the nonlinear echo state network and adaptive neuro fuzzy inference system （ANFIS），the mean absolute error （MAE），root mean square error （RMSE），normalized mean absolute error （NMAE）and normalized root mean square error （NRMSE）of the improved NESN model are significantly reduced，and the prediction accuracy and stability are improved.The accuracy of the wind turbine alignment the wind direction is improved and the mechanical loss of yaw is reduced.

Over the years，machine learning has made some breakthroughs in the insulation defects of gas insulated switchgear（GIS），but the traditional methods have the disadvantages of incomplete information，excessive reliance on artificial feature extraction and low diagnosis rate. In order to solve these problems，a diagnosis method based on deep graph convolutional neural network （DGCN）was proposed. Firstly，a partial discharge （PD） experimental platform was built on a 220 kV real GIS and the partial discharge signals collected by ultra high frequency sensor were converted into frequency domain spectrogram samples by Fourier transform. Then，the spectrogram samples were input into the DGCN，which undergoes graph convolution，coarsening and pooling operations to make the spectrogram structure was clearer and enrich the input information. Finally，the test samples were used to test the DGCN with set parameters. The experimental results show that the proposed method can achieve a recognition rate of 98.77% for GIS fault defects，which is significantly higher than other methods and has good robustness.

Flux-weakening control technology can effectively broaden the speed range of the permanent magnet synchronous machine（PMSM）， therefore， it is of great significance to the research on the flux-weakening control of permanent magnet synchronous machine. Most of flux-weakening modified the flux-weakening current command through a single PI link， and its response speed is slow and difficult to meet the high-performance control requirements， and due to the high coupling and nonlinearity of the flux-weakening loop， PI parameters setting is difficult. In order to solve the above shortage， a method of using a conversion look-up table was proposed to be added to the control system as a feedforward， having accelerated the speed of flux-weakening adjustment.While， the method of gain linearization was adopted， by analyzing the changes of the small signal model at different operating points， it solved the problem that the fixed PI parameters of the flux-weakening loop controller that have been set by experience cannot cope with the change of the operating point， which improved the stability and accuracy of the system. Finally， according to the simulation and experiments， it was verified that the improved method can improve the performance of flux-weakening control.

As the penetration rate of new energy and power electronic equipment increases，the inertia and damping of the power system decrease，and the stability decreases.The grid-forming control technology builds a voltage source to support the stable operation of the large power grid through the converter，so as to play the role of rapid frequency and voltage regulation，increase inertia and short-circuit capacity support，and suppress broadband oscillation，so it has gradually attracted attention. Firstly，the differences between the grid-following converter and the grid-forming converter were compared，and the grid-forming control technology was more suitable for the new power system. Secondly，according to the four different control strategies of the grid-forming converter，the control mechanism of these strategies was analyzed in principle.Finally，by using the same initial parameters and simulating the influence of different control strategies on the power grid frequency after the power grid was disturbed by common disturbances，the superiority of VSG control strategy was verified，and the influence of VSG virtual moment of inertia and damping coefficient on the output frequency response of the system was discussed，which provides a reference for further in-depth research.

During the joint operation of permanent magnet direct drive wind power generation systems，the changes in motor side parameters may cause problems such as reduce control accuracy and stability on the machine side，which in turn affects the DC bus voltage of the converter，leading to power fluctuations on the grid side and posing a threat to the stability of the power system. When using traditional single vector model free predictive control on the machine side，although it effectively improves system robustness，there is still a problem of poor control effect. Therefore，an improved model free predictive control suitable for permanent magnet direct drive wind power generation systems was proposed to maintain system robustness while improving steady-state control performance. Firstly，the machine side and grid side models during the joint operation of the system were analyzed. Secondly，through the method of the extended finite control set，the traditional finite control set was optimized. Combined with the model free predictive control concept of traditional current difference，the proposed method improves the prediction accuracy of the control method and reduces current and power ripple. Finally，the effectiveness and superiority of the proposed method were verified through experiments.

Due to the advantages of faster dynamic performance and smaller torque ripple， high frequency pulse voltage injection （PSVI） is widely used in sensorless control of permanent magnet synchronous motor （PMSM） at low-speed range. Under low switching frequency， the dynamic and steady performance of the traditional high frequency pulse voltage injection method is seriously affected due to the larger system delay， the high-frequency phase update delay caused by the special control structure， and the increase of the phase delay caused by the lower cut-off frequency of the filter. In view of the above problems， compensation schemes for different non-ideal factors were proposed based on the detailed analysis of the influence of the filter phase delay and the delayed phase update of high frequency pulse voltage under low switching frequency on the rotor position estimation. Finally， the influence of the above non-ideal factors and the compensation scheme were simulated and verified. The simulation results show that the proposed solution can effectively reduce the rotor position estimation error.

The integration of large-scale wind power，photovoltaic and other renewable energy resources for the power system will greatly change the power supply characteristics，reducing the inertia of power system and threatening the stability of transient frequency. The virtual synchronous generator（VSG） technology can improve the inertia of renewable energy to promote the stability of transient frequency. In order to quantitatively describe the characteristics of transient frequency of VSG and enhance the control ability of VSG，the transient frequency model of VSG needs to be studied urgently. Based on the VSG control strategy of renewable energy，considering the inertial response，primary frequency regulation of renewable energy，the transient frequency model of VSG was established，and a Matlab/Simulink based simulation model was built to verify the validity of the established analytical model. Finally，the influence of different control parameters on the transient frequency of VSG was analyzed，and the effect of control parameters on the transient frequency of VSG was clarified，giving the guidance for system frequency stabilization.