The AC voltage is very sensitive to the change of reactive power in weak AC systems. If the both control of STATCOM and HVDC in the converter station is not coordinated properly, too fast recovery of DC power after the failure will lead to serious issues such as reactive power imbalance, voltage distortion, repeated commutation failures, which will have a great impact on the stable operation of HVDC. The paper analyzes the problems of HVDC recovery after the failure of weak AC system, proposes a comprehensive optimization method of VDCOL for STATCOM and HVDC system, builds a RTDS simulation model for joint operation of STATCOM and HVDC， and carries out simulation and field experiment verification. The results show that the proposed optimization method can limit the DC current overshoot, reduce the reactive power consumption. Meanwhile the appropriate extension of DC recovery time and the effective cooperation of STATCOM are beneficial for AC voltage recovery and reduce subsequent commutation failures, and have a significant effect on improving DC stable operation in the weak receiving-end.

The DC circuit breaker is the key equipment of the DC system, which can realize the current fault clearing and isolation. The arc extinguishing time of the DC circuit breaker based on the artificial zero-crossing point is uncertain, and the commutation capacitor needs to be replaced with the auxiliary power supply, which reduces the reliability of the circuit breaker and increases the maintenance cost. Hence, a new topology design scheme of the hybrid DC circuit breaker is designed based on photovoltaic charging. In this scheme, the IGBT switch is locked when the DC circuit breaker crosses zero-crossing point and the arc is not extinguished, which accelerates the arc extinguishing, and the photovoltaic energy storage power supply system is designed to precharge the commutation capacitor to ensure the continuous operation of the circuit breaker. At the same time, the topology, breaking sequence and topology parameters of the proposed DC circuit breaker are designed, based on the MATLAB/Simulink experimental platform，the 150 V hybrid DC circuit breaker model is built. The blocking time of IGBT switching devices and the influence of resistance-capacitance diode (RCD) snubber circuit on the breaking of circuit breakers are discussed. On this basis, the breaking processes of the circuit breaker under the normal and short-circuit conditions are analyzed by simulation experiments, the validity of the proposed scheme is verified.

This paper studies the influence of parasitic inductance on overvoltage of medium voltage DC solid state circuit breaker (SSCB) based on the normally-on silicon carbide (SiC) junction field effect transistor (JFET) series structure, and proposes an overvoltage suppression method for SSCB on this basis. Firstly, the topology and operation principle of the SSCB based on the series structure of the normally-on type SiC JFETs are introduced, and the mathematical model of the switching process for the SiC JFET series structure considering the parasitic inductance of the complete loop is established. Secondly, MATLAB software is used to analytically calculates the mathematical model, and the mechanism of overvoltage generated by parasitic inductance during SSCB switching to SiC JFET devices in series operation is revealed. At the same time, the simulation results of PSPICE verify the correctness of the theoretical analysis. Finally, single gate drive and snubber circuit are designed for SSCB overvoltage suppression, and the SSCB experimental prototype verifies the effectiveness of the proposed method.

The actual large-scale power system at the provincial level or above has numerous nodes, and the input feature space of deep learning models used for transient stability assessment of power grids faces the curse of dimensionality. The high training cost and difficulty in ensuring generalization ability have become a bottleneck for the application of such methods in practical large-scale systems. A graph attention deep learning transient power angle stability evaluation model is proposed to address this issue, which is oriented towards stability control verification and utilizes network equivalence for graph dimensionality reduction. Firstly, a whole grid generators diagram is established, and node-similarity is constructed based on key parameters such as rotor inertias that determine transient power angle stability, and node similarity is used to improve the edge weights of the generator diagram. As a domain knowledge embedding method, the dynamic equivalence approach of the power grid is used for reference, the network parts outside the area involved in the study of stability control systems are partitioned based on generators diagrams and hierarchical clustering algorithms. The formed partitions are corresponding to equivalent nodes to form a reduced dimensional generators diagram, and a mapping of the original graph to the reduced dimensional graph nodes and edge weight parameters is established to achieve dimensionality reduction of the original input space. Finally, a graph attention deep learning model is established using the dimensionality reduction graph as input to achieve transient power angle stability evaluation of complex networks. The effectiveness and accuracy of the model are verified through comparative analysis on a practical stability control system example in China Southern Power Grid.

The possible characteristics of out-of-step oscillation of multi-section and multi-channel interconnection power grid are summarized and analyzed, such as the selectivity of multi-section out-of-step, the different out-of-step time of different channels on the same section and the lack of obvious out-of-step characteristics of some channels on the same section. In response to the problem that the existing local out-of-step splitting schemes can′t adapt to the above characteristics, a wide area out-of-step splitting control system method based on the action permission command of local splitting device and constructed through remote communication is adopted. A wide area out-of-step splitting control system of China Southern Power Grid is designed and established with the main principles of considering solving the problem of the existing local out-of-step splitting scheme, the coordination with the existing local out-of-step splitting function, system error prevention design and the control strategy processing logic under abnormal circumstances. The real-time digital simulator （RTDS) simulation test results show that the system can respond correctly, selectively, quickly, and reliably take established measures to split the related sections at the same time, effectively solving the limitation of the existing local out-of-step splitting scheme and meeting the operational needs of actual power system.

Aiming at the prominent wind abandonment problem in "Three Norths" areas, an electrothermal economic dispatch scheme is proposed to realize the wind consumption using the CHP steam turbine bypass compensation heating. The dispatch model takes into account the stepped carbon trading mechanism, and two electrothermal decoupling schemes for heat supply compensation, namely, the whole bypass and the HP-LP bypass are considered. The coal consumption calculation model of bypass heating is established based on the heat quality method. Effects of the bypass heating during the peak period of wind power generation at night on coal consumption and wind curtailment are studied, and the electrothermal operation characteristics of thermal power units are analyzed. The calculation example results show that the dispatch scheme of the whole bypass heating can effectively achieve the electrothermal decoupling, which brings the best low-carbon economic electrothermal distribution results, and the CHP steam turbine has good electrothermal operation characteristics.

With the integration and development of smart grid and 5G communication technology, more and more intelligent terminals are applied to the smart grid system. Aiming at the problem of diversion processing of massive power services, a service scheduling and resource allocation scheme considering the priority of power service is proposed. Firstly, the edge computing processing architecture based on software defined network for smart grid is introduced, and the service processing models are established. Secondly, the service scheduling mechanism based on preemptive priority queuing is elaborated, the mathematical models of service offloading revenue and offloading cost are established with overall system revenue maximization as the objective function. Based on the offloading effectiveness of power service, the resource allocation threshold of each priority service in the mobile edge computing (MEC) server is obtained and taken as the constraint condition. Thirdly, the improved genetic algorithm (IGA) is used to solve the optimal offloading and resource allocation decisions. Finally, the simulation results demonstrate that IGA outperforms other comparative algorithms in terms of convergence speed and individual selection, and the proposed scheme reduced the average processing time, energy consumption, and average processing time of high priority services by 69.2%, 67.7%, and 73% compared to other methods, respectively. The system revenue is improved by 119% compared to other methods.

The ultrasonic testing method is widely used in the detection and positioning of insulation status of power equipment. The directional MUSIC （Dir-MUSIC） direction-finding algorithm based on direction-finding ultrasonic array signal strength information is proposed to overcome the shortcomings of MUSIC algorithm for partial discharge ultrasonic direction-finding, such as high sampling rate requirements and large computational complexity. On the basis of elaborating on the theoretical model and application conditions of the algorithm, the paper simulates and studies the direction-finding accuracy of the Dir-MUSIC algorithm under different gain pattern main lobe width and signal-to-noise ratio conditions for a uniform disc ultrasonic array. The simulation results show that the 8-element array has the highest direction-finding accuracy at a signal-to-noise ratio of -5 dB and a main lobe width of 90 °~120 °, with a root mean square error of less than 2 °. Finally, a direction-finding experiment is conducted based on the developed MEMS direction-finding ultrasound array. The results show that the root mean square error of direction-finding for the 8-element disc ultrasound array is the minimum of 2.76 °, and the standard deviation of direction-finding is the minimum of 2.72 °, which verify the effectiveness and accuracy of the Dir-MUSIC algorithm.

To analyze the vibration characteristics of transformer windings with high accuracy, a solution approach combining the axial dynamic model of windings with the two-stage fourth-order Gauss implicit Runge Kutta （GRK24） method is developed. The simplified model of "mass-spring-damping" is adopted for the axial dynamics modeling of windings, and the vibration equation is solved by the matrix form GRK24 （GRK24-M） method in which the displacement, velocity and acceleration solutions can reach fourth-order accuracy simultaneously. The windings structure of a 50 kVA transformer is taken as an example. The effectiveness of the proposed approach solving the windings vibration responses under the impact load and the electromagnetic force load is verified. The superiority of the proposed GRK24-M method is also illustrated by comparing the solution results with the Newmark method and the HHT-α method.

The security problem of the global positioning system （GPS） deception signal transmitted by unmanned aerial vehicle （UAV） to the automatic system in the substation is not clear. This paper proposes a method to evaluate the interference of GPS deception signal to the time synchronization device of substation, and theoretically analyzes the influence of GPS deception signal on the time synchronization device of substation. The spatial distribution of deception signal power intensity and the electric field intensity of unmanned aerial vehicle decoy equipment under different transmitting power are measured. The experiment verifies the influence of GPS deception signal on the quality of the satellite received by the time synchronization device and the accuracy of time service. The influence laws of the transmitting power of the unmanned aerial vehicle decoy equipment and the distance between it and the GPS/BeiDou navigation satellite system （BDS） receiving antenna of the time synchronization device on the time service accuracy are analyzed. The results show that the GPS decoy signal will not make the time synchronization device enter the punctual state when the antenna of the time synchronization device is 10 m away from the unmanned aerial vehicle decoy device. The antenna of the time synchronization device is more than 60 m away from the unmanned aerial vehicle decoy equipment, and the impact of GPS deception signal on the time synchronization device can be ignored, which provides reference for safe and stable operation of substation.

Transformer oil de-composition into combustible gas combustion has a high risk. The new environmental protection fire extinguishing agent C₆F₁₂O has poor reignition resistance and will produce harmful product HF. Therefore, this paper studies the effects of N₂ or CO₂ addition on the fire extinguishing performance of C₆F₁₂O extinguishing the flame with transformer oil discharges gas and the harmful product HF. Chemical dynamics method is used to calculate the effects of C₆F₁₂O mixture on adiabatic combustion temperature and active free radical concentration of transformer oil de-composition gas flame under different equivalent ratios and mixing ratios. The variation law of HF concentration and generation path are analyzed. Theoretical calculation results show that N₂ or CO₂ and C₆F₁₂O have good physicochemical collaborative fire extinguishing effect and can inhibit the formation of harmful product HF. The research results of this paper provide theoretical basis and data reference for the application of C₆F₁₂O fire extinguishing agent in electrical fire.

Aiming at the increasingly urgent power supply demand of users and the power quality problem caused by the randomness and volatility of photovoltaic output and load power consumption，a bi-level optimization model of photovoltaic-storage microgrid considering power supply demand is established, and the power quality of the planned microgrid is evaluated. The model takes the minimum standard deviation of equivalent net load and abandonment photovoltaic rate as the upper optimization objective, and the minimum active power loss of microgrid as the lower optimization objective. The energy storage is used to meet the power demand of the guaranteed power supply load as an additional constraint condition, and the genetic algorithm is used to solve the model. Then, the source side, load side and grid side indexes are designed, using the combination weighting method to determine the weight of each evaluation index to achieve the comprehensive evaluation of microgrid power quality. Finally, the effectiveness of the proposed model is verified by the improved IEEE-33 system. The results show that the power quality of microgrid can be improved after considering the power supply demand, and within a certain range, the larger the power supply load, the better the power quality and user satisfaction of the microgrid.

Energy storage possesses flexible power regulation and energy time-shifting capabilities, and its coordination and optimization with multiple types of regulatory resources in the system is expected to provide strong support for the low-carbon, economic, and reliable operation of new power systems. Research on the benefits of low-carbon, economic, and reliable performance of high proportion renewable energy bases for energy storage improvement is conducted. Firstly, a carbon emission measurement model for power generation in high proportion renewable energy bases is established, and evaluation indicators for low-carbon, economic, and reliable energy utilization in high proportion renewable energy bases are proposed. Secondly, a joint operation simulation model is established for energy storage and high proportion renewable energy bases, providing decision-making basis for the optimized operation of renewable energy bases with high proportion of energy storage. Finally, a fine-grained annual time scale operation simulation will be conducted for the Mengxi regional power grid to analyze the benefits of improving the low-carbon, economic, and reliable operation performance of the system under different energy storage capacity ratios. The calculation results have verified the effectiveness of the proposed operational simulation model and pointed out that as the energy storage capacity increases, the marginal benefits of improving the low-carbon, economic, and reliable performance of the system gradually decrease. Formulation optimization of energy storage configuration capacity is beneficial for maximizing its investment and operational benefits.

With a large number of new energy being connected to the distribution network, the uncertainty and volatility of new energy output have brought great challenges to the distribution network planning. The impact of uncertainty and volatility of new energy on planning results can be reduced by comprehensively considering the source-network-load-storage. Based on this, a joint planning method of source-network-load-storage in distribution network considering dynamic reconfiguration and intelligent soft open point is proposed. First of all, according to the idea of density peak clustering, an improved affinity propagation clustering algorithm based on the density peak is proposed to cluster the combined scenes of wind power, photovoltaic and load to obtain the typical daily curve. Then, taking the minimum of the total planning cost as the objective function, a joint planning model of source-network-load-storage in distribution network considering dynamic reconfiguration and intelligent soft open point is established. Based on the second-order cone theory, the original non convex nonlinear planning model is transformed into a mixed integer second-order cone planning model. Finally, the validity of the proposed model and method is verified by simulation on the Portugal 54 numerical example.

An interphase fault location method based on random forest algorithm is proposed targeting the impact of distribution line parameter errors on the results of interphase fault distance measurement. Firstly, four fundamental double-side fault location algorithms are selected, and the relationship among parameter error, measurement error, and ranging results is analyzed. The result show that it is difficult to describe the relationship by a linear and analytical expression. Secondly, a random forest model is established, and a nonlinear model is constructed with basic algorithm ranging results as input and precise fault location as output to describe the non-linear relationship among parameter error, measurement error, and ranging results. Therefore interphase fault distance measurement is completed. Finally, the digital simulation results show that the proposed model has high location accuracy and is not affected by parameter errors, fault location, initial fault angle and other factors. Also, the model is suitable for distribution network with high penetration distributed generation and has generalization ability and practical value has practical value.

There is a great demand for strain clamp in transmission lines, but the strain clamp failure occurs frequently, which not only brings economic losses, but also seriously threatens the reliable operation of power grid. In order to explain the failure mechanism of hydraulic strain clamp and improve the safety of transmission lines, scholars at home and abroad have carried out a lot of research. Overheating and fracture engineering failure types of hydraulic strain clamp are reviewed from the aspect of failure accidents, the failure mechanisms of two engineering failure types are reviewed from the aspects of engineering analysis and mechanism research, and the influence of process parameters on failure forms is reviewed from the aspects of crimping process research. The research status of quality testing of strain clamp is summarized from the perspectives of routine testing and nondestructive testing. Finally, based on the current research status of partial qualitative analysis on the engineering failure of the strain clamp, the prospect of further research on the failure mechanism of the strain clamp is proposed.

Automatic power line segmentation is an important prerequisite for the safe operation of intelligent inspection platforms. However, power line segmentation is a small target segmentation problem in complex backgrounds and multiple climatic environments, which is highly prone to encounter false or missed detections. In order to improve the robustness and accuracy of power line segmentation, an end-to-end segmentation model based on hierarchical attention fusion is proposed in combination with an encoder-decoder framework. The model proposes a reduced-dimensional residual convolution unit that increases the network depth while significantly reducing the network parameters, making it easier to deploy in embedded devices, enabling the model to capture global information and emphasize the target regions of powerlines, a chain-based hierarchical attention fusion module is designed for multi-scale feature fusion to address the category imbalance problem. To improve the model's attention to the unique line prior features of power lines, the line prior loss function is combined with the Focal loss function and Dice loss function to form a joint loss function to further improve the accuracy of power line segmentation. The experimental results show that the depth of the proposed model network increases to about 2.8 times that of the base network, while the number of parameters is only about 1/3 of the original one. Robust segmentation of power lines can be achieved for both regular weather and foggy weather aerial images. The proposed model can be applied to the field of power inspection, making the inspection more intelligent and efficient.