Power grid fault emergency rescue method

Abstract

The invention relates to the technical field of power grid fault repair, in particular to a power grid fault emergency rescue method. The method comprises the following steps: acquiring topological structure data, key equipment parameters and real-time operation state data of a power grid; constructing a power grid digital twin model according to the power grid topological structure data, the key equipment parameters and the real-time operation state data, and performing real-time state evaluation to obtain fault risk early warning data and real-time fault condition data; when the real-time fault condition data shows that a fault occurs, performing fault point positioning and fault type determination based on deep learning according to the fault risk early warning data and the real-time operation state data to obtain fault diagnosis data; and constructing a fault influence range model according to the fault diagnosis data and a preset FACTS and HVDC characteristic model. According to the method, deep learning, a graph theory and a multi-objective optimization technology are fused, and intelligent fault identification, influence range modeling and accurate emergency disposal scheme optimization are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid fault repair, and in particular to a power grid fault emergency rescue method. Background Art

[0002] As a key infrastructure for energy supply in modern society, the safe and reliable operation of the power grid is of vital importance. However, due to the large scale and complex structure of the power grid system, coupled with the influence of various natural and human factors, the occurrence of fault events is inevitable. Power grid failures may cause power outages in some or large areas, which will have a serious impact on social production and life. Therefore, timely and effective handling of power grid failures and rapid restoration of power supply are the top priorities for the operation of the power system. Among them, power grid failures include equipment failures: mainly mechanical failures, insulation failures, overheating, etc. of key equipment such as transformers, circuit breakers, and busbars, which may cause equipment damage, tripping or burning; line failures: due to lightning strikes, broken branches, wind, rain, snow and ice disasters, etc., transmission lines trip, insulator string damage, wire breakage or tower tilting; relay protection misoperation: the relay protection system misjudges or fails to judge the fault, causing non-fault lines or equipment to be mistakenly cut off, etc. If these faults are not handled properly, they are very likely to spread and expand, leading to a larger range of power outages, bringing catastrophic effects to the power grid system.

[0003] Traditional emergency rescue methods for power grid faults often have the following problems: the gradual introduction of more new intelligent devices and new technologies into the power grid, such as FACTS, HVDC, distributed power sources, etc., brings new challenges to fault diagnosis and emergency dispatch; traditional emergency plans and black start plans are often static and experience-driven, lacking comprehensive assessment and optimization of complex dynamic environments, which can easily lead to waste of resources or improper disposal. Summary of the invention

[0004] Based on this, it is necessary for the present invention to provide a power grid fault emergency rescue method to solve at least one of the above technical problems.

[0005] To achieve the above purpose, a power grid fault emergency rescue method includes the following steps: Step S1: Obtain power grid topology data, key equipment parameters, and real-time operating status data; construct a power grid digital twin model based on the power grid topology data, key equipment parameters, and real-time operating status data, and perform real-time status evaluation to obtain fault risk warning data and real-time fault condition data; Step S2: When the real-time fault condition data indicates that a fault has occurred, the fault point is located and the fault type is determined based on deep learning according to the fault risk warning data and the real-time operation status data, thereby obtaining fault diagnosis data; a fault impact range model is constructed according to the fault diagnosis data and the preset FACTS and HVDC characteristic models; Step S3: Generate multiple sets of emergency processing sub-plans based on the multi-objective optimization algorithm according to the fault impact range model to obtain an emergency sub-plan set, perform simulation verification on the emergency sub-plan set, conduct feasibility and impact assessment on each sub-plan, and combine the sub-plans to obtain optimized emergency plan data; Step S4: obtaining a black start capability model of the current power grid, formulating a restoration strategy for the power grid system according to the optimized emergency plan data and the black start capability model, thereby generating black start path planning data; Step S5: Formulate a load recovery strategy based on load characteristics according to the black start path planning data, and use an adaptive control algorithm to adjust the load access sequence and power in real time, thereby obtaining load recovery control data; Step S6: Design a multi-level control system architecture based on the load recovery control data, black start path planning data, and optimized emergency plan data, and use edge computing technology to control the device nodes to obtain a system recovery execution plan.

[0006] The present invention obtains the grid topology data, key equipment parameters and real-time operation status data, ensuring the accuracy and integrity of the grid digital twin model; through real-time monitoring and analysis, the potential fault risks of the grid can be discovered in time, the early warning capability is improved, and the probability of fault occurrence is reduced; based on the fault risk early warning data generated by real-time status evaluation, preventive measures can be taken in advance to avoid the occurrence of large-scale faults. Using the deep learning algorithm, the fault point and fault type can be quickly located after the fault occurs, greatly shortening the fault location time; based on real-time data and preset models, the cause of the fault can be accurately diagnosed, providing reliable data support for subsequent emergency treatment; by constructing a fault impact range model, the impact of the fault on each part of the grid can be comprehensively evaluated, providing a scientific basis for the formulation of emergency plans. Based on the multi-objective optimization algorithm, multiple sets of emergency treatment sub-plans can be generated to ensure the comprehensiveness and flexibility of emergency treatment; simulation verification of the emergency sub-plans can evaluate the feasibility and impact of each plan in advance to ensure that the selected plan is the best; by combining and optimizing the sub-plans, the most suitable emergency treatment plan for the current fault situation can be formed to improve the emergency treatment effect. By acquiring and analyzing the black start capability model of the power grid, a scientific and reasonable power grid recovery strategy can be formulated to ensure the rapid recovery of the system; according to the black start path planning data generated by the optimized emergency plan data and the black start capability model, it can ensure that the power grid operation is restored in the shortest time and reduce the power outage time and scope. The load recovery strategy formulated based on load characteristics can ensure that no new load shock will be caused during the recovery process, and improve the safety and stability of the recovery process; the use of adaptive control algorithms to adjust the load access sequence and power in real time can dynamically respond to load changes and ensure the smooth progress of the recovery process. Through the design of a multi-level control system architecture, comprehensive control from device nodes to the system level can be achieved to ensure the coordination and integrity of the recovery process; the use of edge computing technology to control and protect device nodes in real time can improve the control response speed and accuracy and ensure the safe operation of key equipment; integrating load recovery control data, black start path planning data and optimized emergency plan data to form a system recovery execution plan can achieve comprehensive and rapid recovery of the power grid and improve the reliability and safety of the power grid. Through the implementation of the above steps, not only can the efficiency and effect of emergency handling of power grid faults be improved, but also the impact of faults on the power grid and users can be effectively reduced, and the overall operation reliability and safety of the power grid can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments thereof made with reference to the following drawings: Figure 1 A schematic diagram of the steps of the power grid fault emergency rescue method of the present invention; Figure 2 for Figure 1 Detailed step flow diagram of step S1; Figure 3 for Figure 1 Detailed step flow chart of step S2 in FIG. DETAILED DESCRIPTION

[0008] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0009] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0010] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0011] To achieve this, please refer to Figures 1 to 3 The present invention provides a method for emergency rescue of power grid failure, the method comprising the following steps: Step S1: Obtain power grid topology data, key equipment parameters, and real-time operating status data; construct a power grid digital twin model based on the power grid topology data, key equipment parameters, and real-time operating status data, and perform real-time status evaluation to obtain fault risk warning data and real-time fault condition data; Step S2: When the real-time fault condition data indicates that a fault has occurred, the fault point is located and the fault type is determined based on deep learning according to the fault risk warning data and the real-time operation status data, thereby obtaining fault diagnosis data; a fault impact range model is constructed according to the fault diagnosis data and the preset FACTS and HVDC characteristic models; Step S3: Generate multiple sets of emergency processing sub-plans based on the multi-objective optimization algorithm according to the fault impact range model to obtain an emergency sub-plan set, perform simulation verification on the emergency sub-plan set, conduct feasibility and impact assessment on each sub-plan, and combine the sub-plans to obtain optimized emergency plan data; Step S4: obtaining a black start capability model of the current power grid, formulating a restoration strategy for the power grid system according to the optimized emergency plan data and the black start capability model, thereby generating black start path planning data; Step S5: Formulate a load recovery strategy based on load characteristics according to the black start path planning data, and use an adaptive control algorithm to adjust the load access sequence and power in real time, thereby obtaining load recovery control data; Step S6: Design a multi-level control system architecture based on the load recovery control data, black start path planning data, and optimized emergency plan data, and use edge computing technology to control the device nodes to obtain a system recovery execution plan.

[0012] In the embodiment of the present invention, reference Figure 1 The above is a schematic flow chart of the steps of a power grid fault emergency rescue method of the present invention. In this example, the power grid fault emergency rescue method includes the following steps: Step S1: Obtain power grid topology data, key equipment parameters, and real-time operating status data; construct a power grid digital twin model based on the power grid topology data, key equipment parameters, and real-time operating status data, and perform real-time status evaluation to obtain fault risk warning data and real-time fault condition data; The embodiment of the present invention obtains the physical connection relationship of the substation, transmission line, and distribution network in the power grid, which usually includes a node-branch structure, which can be obtained through a SCADA system or a power grid management system; collects parameters of major equipment such as transformers, circuit breakers, generators, etc., including rated capacity, impedance, rated current, etc., which can be obtained from data tables or equipment maintenance records provided by equipment manufacturers; real-time monitoring of the voltage, current, power, frequency and other data of the power grid is usually provided by a monitoring system (such as a SCADA system). The network topology model is constructed using the power grid topology data to form a node-branch structure, each node represents a key device or contact in the power grid, and each branch represents a cable or line connecting these nodes; attributes are added to each node and branch according to the key equipment parameters, and the electrical characteristics of the equipment (such as impedance, power capacity) are integrated into the model; the model is dynamically assigned a state according to the real-time operating status data, including voltage, current, power, etc., to ensure that the model reflects the actual operating status of the power grid; power system analysis software is used to perform flow calculations, stability analysis, etc., to generate real-time status data, thereby obtaining fault risk warning data and real-time fault condition data.

[0013] Step S2: When the real-time fault condition data indicates that a fault has occurred, the fault point is located and the fault type is determined based on deep learning according to the fault risk warning data and the real-time operation status data, thereby obtaining fault diagnosis data; a fault impact range model is constructed according to the fault diagnosis data and the preset FACTS and HVDC characteristic models; When a fault is detected, the embodiment of the present invention extracts high-resolution fault waveform data before and after the fault occurs; uses a filtering algorithm (such as a low-pass filter) to remove noise in the waveform data; uses a pre-trained convolutional neural network (CNN) to locate the fault point of the processed waveform data and identify the specific location of the fault; classifies the fault waveform data through a support vector machine (SVM) to determine the type of fault, such as line short circuit, equipment failure, etc. The preset FACTS (Flexible AC Transmission System) and HVDC (High Voltage Direct Current Transmission System) characteristic models are used to analyze the dynamic response of the fault to the power grid; based on the fault diagnosis data and characteristic model, a fault impact range model is constructed to evaluate the degree and range of the impact of the fault on various parts of the power grid.

[0014] Step S3: Generate multiple sets of emergency processing sub-plans based on the multi-objective optimization algorithm according to the fault impact range model to obtain an emergency sub-plan set, perform simulation verification on the emergency sub-plan set, conduct feasibility and impact assessment on each sub-plan, and combine the sub-plans to obtain optimized emergency plan data; The embodiment of the present invention generates multiple emergency response plans based on the fault impact range model, such as dynamic reconstruction of the power grid topology, reconfiguration of power supply, isolation of fault area, load transfer, etc.; the generated emergency response plans are integrated into a set of emergency sub-plans. Each emergency sub-plan is simulated and tested using the power grid digital twin model to verify its performance under different fault conditions; the simulation results are evaluated, the recovery effect, economic cost, system stability, etc. of each plan are analyzed, and a multi-objective optimization model is constructed; the sub-plans are combined according to the evaluation results, the optimal emergency response plan is determined, and optimized emergency plan data is generated.

[0015] Step S4: obtaining a black start capability model of the current power grid, formulating a restoration strategy for the power grid system according to the optimized emergency plan data and the black start capability model, thereby generating black start path planning data; The embodiment of the present invention collects black start capability data of each generator set in the power grid, including the starting capability, starting time, output power, etc. of the generator set. A black start capability model is constructed based on the black start resource list and performance information to evaluate the starting and recovery capabilities of each generator set. The black start capability model is used to formulate a recovery strategy for the power grid system, including the starting sequence, recovery steps, etc., and generate black start path planning data.

[0016] Step S5: Formulate a load recovery strategy based on load characteristics according to the black start path planning data, and use an adaptive control algorithm to adjust the load access sequence and power in real time, thereby obtaining load recovery control data; The embodiment of the present invention evaluates the priority of loads according to load characteristic data and black start path planning data, formulates a preliminary load recovery strategy; adjusts the load recovery strategy to ensure that the most critical loads can be restored first during the recovery process; monitors the load recovery status of the power grid system in real time and obtains real-time load data; uses an adaptive control algorithm to adjust the load access sequence and power according to real-time data to ensure the stability and safety of the load recovery process, and generates load recovery control data.

[0017] Step S6: Design a multi-level control system architecture based on the load recovery control data, black start path planning data, and optimized emergency plan data, and use edge computing technology to control the device nodes to obtain a system recovery execution plan.

[0018] The embodiment of the present invention designs a multi-level control system architecture, including central control, regional control and local control layers, to meet the control requirements at different levels; integrates the optimized emergency plan data, black start path planning data and load recovery control data into the control system. Deploy edge computing devices on key power grid equipment nodes to achieve local data processing and real-time control; use edge computing technology to perform real-time control and protection of equipment, ensure the stability and security of the system during the recovery process, and generate a system recovery execution plan.

[0019] Preferably, step S1 comprises the following steps: Step S11: collecting substations, transmission lines, distribution networks and their physical connection relationships in the power grid, and integrating the devices and their connection relationships to obtain power grid topology data; Step S12: collecting equipment parameters of transformers, transmission lines, circuit breakers, capacitors and reactors, distribution transformer battery energy storage systems, and generators to obtain key equipment parameters; Step S13: Real-time monitoring of voltage, current, power, frequency and power flow information is performed through the substation auxiliary service system to obtain real-time operation status data; Step S14: standardizing the grid topology data, key equipment parameters and real-time operation status data, and removing data with abnormal formats, thereby obtaining comprehensive grid data; Step S15: constructing a power grid digital twin model based on comprehensive power grid data; Step S16: Perform real-time status evaluation on the power grid digital twin model to obtain fault risk warning data; Step S17: Perform feature comparison and abnormality identification based on a preset fault feature library and real-time operating status data, thereby obtaining real-time fault condition data.

[0020] As an embodiment of the present invention, refer to Figure 2 As shown, Figure 1 Detailed step flow diagram of step S1 in the embodiment of the present invention, step S1 includes the following steps: Step S11: collecting substations, transmission lines, distribution networks and their physical connection relationships in the power grid, and integrating the devices and their connection relationships to obtain power grid topology data; The embodiment of the present invention obtains the location information, equipment configuration and connection relationship of the substation through the power system monitoring system (such as SCADA), obtains the starting and ending points, line type, capacity and other data of the transmission line, and collects information such as distribution transformers and branch lines in the distribution network. The physical connection relationship of the substation, transmission line and distribution network is integrated to form a complete power grid topology map, which is visualized and processed using geographic information system (GIS) software; a node-branch model is created, in which the node represents the equipment (such as the substation, generator), and the branch represents the cable or line connection, and the power grid topology structure data is established.

[0021] Step S12: collecting equipment parameters of transformers, transmission lines, circuit breakers, capacitors and reactors, distribution transformer battery energy storage systems, and generators to obtain key equipment parameters; The embodiment of the present invention collects equipment parameters, including transformer: obtaining the rated capacity, impedance, transformation ratio and other parameters of the transformer; transmission line: collecting the resistance, reactance, conductivity and other electrical characteristics of the line; circuit breaker: recording the rated current, breaking capacity and other characteristics of the circuit breaker; capacitor and reactor: obtaining their rated capacity, operating voltage, access mode and other characteristics; distribution transformer, battery energy storage system: obtaining parameters such as capacity, charge and discharge efficiency, and charging state; generator: recording the rated power, speed, load capacity and other characteristics of the generator. The collected data is formatted and entered into the database for subsequent use; the data is checked for accuracy and completeness to ensure that it is correct.

[0022] Step S13: Real-time monitoring of voltage, current, power, frequency and power flow information is performed through the substation auxiliary service system to obtain real-time operation status data; The embodiment of the present invention monitors the operation status of the power grid in real time, including voltage, current, power and frequency, through the substation auxiliary service system (such as the SCADA system); collects real-time data of the power flow in the power grid, including power flow direction and power loss, etc. The monitoring data is transmitted to the central database in real time through the data acquisition system; the data is stored in an efficient database to support subsequent analysis and model building.

[0023] Step S14: standardizing the grid topology data, key equipment parameters and real-time operation status data, and removing data with abnormal formats, thereby obtaining comprehensive grid data; The embodiment of the present invention converts power grid data from different sources into a unified standard format for unified processing; unifies data units (such as unifying voltage into volts and current into amperes). Detects and removes outliers and noise in the data, and uses statistical methods such as mean and standard deviation to detect abnormal data; uses interpolation methods (such as linear interpolation and spline interpolation) to process missing values ​​in the data to ensure data integrity. Integrates power grid topology data, key equipment parameters, and real-time operating status data to generate a comprehensive power grid data set.

[0024] Step S15: constructing a power grid digital twin model based on comprehensive power grid data; The embodiment of the present invention establishes a network topology model based on the power grid topology structure data, which includes parameter integration: integrating key equipment parameters into the node-branch model, and updating the electrical characteristics of the nodes and branches; state assignment: assigning values ​​to each node and branch in the model according to the real-time operating status data, including voltage, current, power, etc.; model verification: using historical data to verify the accuracy of the model to ensure that the model can accurately reflect the actual state of the power grid; simulation test: performing model simulation test to ensure that the model can simulate the behavior of the power grid under various operating conditions.

[0025] Step S16: Perform real-time status evaluation on the power grid digital twin model to obtain fault risk warning data; The embodiment of the present invention uses a digital twin model of a power grid to perform power flow calculations to evaluate the voltage, current, and power of each node in the power grid; performs power grid stability analysis to evaluate the stability of the power grid in its current state, including static stability and dynamic stability; and generates fault risk warning data based on the state assessment results to identify possible risk areas and equipment.

[0026] Step S17: Perform feature comparison and abnormality identification based on a preset fault feature library and real-time operating status data, thereby obtaining real-time fault condition data.

[0027] The embodiment of the present invention acquires and maintains a feature library including different fault modes, and these feature modes may include voltage drop, frequency fluctuation, etc. The normal characteristics of the power grid, such as voltage level, load fluctuation, etc., are extracted from the real-time operation status data; compared with the fault feature library, whether there are abnormal features, and abnormality detection is performed using pattern recognition technology. According to the key features in the real-time operation data, the patterns in the fault feature library are matched to confirm whether a fault has occurred; and real-time fault condition data is generated to identify whether a fault has occurred and its specific type.

[0028] The present invention collects data on substations, transmission lines, distribution networks and their physical connections in the power grid to ensure the integrity and accuracy of the power grid topology data; through the integration of devices and their connections, the overall structure of the power grid can be intuitively displayed, which helps to understand the operation mode of the power grid and the relationship between each device; through standardized collection and integration, the consistency of the power grid topology data in various systems and applications is ensured, and the reliability and accuracy of the data are improved. Parameters of various key equipment in the power grid (transformers, transmission lines, circuit breakers, capacitors and reactors, distribution transformers, battery energy storage systems, generators) are collected to ensure that the equipment data is comprehensive and accurate; accurate equipment parameter data can be used to better evaluate the performance and status of each device, which is helpful for equipment operation maintenance and fault prevention; reliable data support is provided for the subsequent construction of digital twin models and real-time status evaluation to ensure the accuracy of the model. Through the substation auxiliary service system, voltage, current, power, frequency and flow information are monitored in real time to ensure that the operation status of the power grid can be obtained in time; real-time monitoring can capture the dynamic changes in the operation of the power grid, which helps to discover and respond to abnormal situations in time; comprehensive operation status data ensures the integrity of the comprehensive data of the power grid, and provides a comprehensive data basis for the construction and status evaluation of the digital twin model. Standardization of power grid topology data, key equipment parameters and real-time operation status data ensures the consistency of data format and improves data processing efficiency; removes data with abnormal formats to ensure the accuracy and reliability of data and avoid the impact of abnormal data on subsequent analysis and model construction; through data cleaning and standardization, the quality of comprehensive data of the power grid is improved, providing high-quality data support for subsequent analysis and application. The digital twin model of the power grid is constructed based on the comprehensive data of the power grid to ensure that the model has high accuracy and high reliability; the digital twin model can truly simulate the operation status and dynamic changes of the power grid, and provide a powerful support tool for operation analysis and decision-making; the digital twin model built based on comprehensive data has good scalability and can be dynamically adjusted and optimized as the data is updated and increased. Real-time status evaluation of the digital twin model of the power grid can timely discover potential risks and abnormal conditions in the operation of the power grid; generate fault risk warning data, which can identify and warn fault risks in advance, reduce the possibility and impact of faults; real-time status evaluation provides important decision-making support for power grid operation and management, and improves the operational safety and reliability of the power grid. According to the preset fault feature library and real-time operation status data, feature comparison and abnormality identification can quickly locate the fault point and fault type; real-time fault situation data provides accurate information support for fault handling, which helps to quickly respond to and solve faults, reduce power outage time and impact; through the combination of fault feature library and real-time data, intelligent fault identification and diagnosis are realized, which improves the efficiency and intelligence level of power grid operation and maintenance.Through the implementation of the above steps, not only the efficiency of power grid data collection and processing is improved, but also the operation monitoring and fault warning capabilities of the power grid are improved, thereby effectively ensuring the safety and reliability of the power grid.

[0029] Preferably, step S15 comprises the following steps: A digital twin model of the power grid is constructed based on the comprehensive data of the power grid. The model construction specifically includes establishing a node-branch model representing the network structure of the power grid and the flow transmission path according to the power grid topology data, adding device nodes and integrating parameters to the node-branch model according to key equipment parameters, and assigning state quantities to each node and branch according to real-time operating status data.

[0030] The embodiment of the present invention extracts the topology of the power grid from the data of the substation, transmission line, and distribution network, including the physical connection relationship between the substation, line, switchgear, etc.; usually GIS (geographic information system) data or network diagram data, including node (substation, generator, load point) and branch (transmission line, distribution line) information. Each substation, generator, load point, etc. is defined as a node in the power grid, and the node includes its location, type, and basic parameters (such as rated voltage, power, etc.); the branch represents the line or cable connecting two nodes in the power grid, and the branch includes the electrical characteristics of the line (such as resistance, reactance) and capacity information; a node-branch model is established using a graph theory method, in which the node is the vertex of the graph and the branch is the edge, which can be implemented using a power system modeling tool (such as PowerWorld, DIgSILENT); a power flow transmission path in the network is established, that is, the path for power to reach the load point from the generator through multiple branches, to ensure that the model can simulate the flow of power. Obtain detailed parameter data (such as rated capacity, impedance, breaking capacity, etc.) of transformers, circuit breakers, capacitors, reactors, distribution transformers, battery energy storage systems, and generators from the equipment parameter acquisition system; save the parameters in a structured data format (such as CSV files, database records). Integrate equipment parameters into the node-branch model. For example, when adding a transformer node, its parameters such as rated capacity and impedance need to be entered into the model. Add equipment characteristics to the nodes to ensure that each node and branch in the model reflects the electrical characteristics of the actual equipment. For example, the transformer node needs to record the transformation ratio and impedance data, and the circuit breaker node needs to record its rated current and breaking capacity. Integrate electrical characteristics (such as resistance, reactance, capacity, etc.) into each node and branch of the model, which includes adjusting the electrical model of each node to ensure that they accurately reflect the operating characteristics of the actual equipment. The voltage, current, power, frequency and other data in the power grid are monitored in real time through the substation auxiliary service system (such as the SCADA system) to ensure that the real-time data can be transmitted to the model, usually through a real-time data interface (such as an API); the state quantity of each node is assigned according to the real-time operating status data, for example, the real-time voltage and current data will be assigned to the substation and generator nodes; the state quantity of the branch is assigned, including real-time flow data (power flow), current and voltage drop. The current and power flow data of the branch will directly affect the electrical calculations in the model; the state quantity in the model is updated in real time to ensure that the model can reflect the actual operating status of the current power grid. Use a real-time data update algorithm (such as an interpolation algorithm) to process the real-time data so that it is consistent with the node and branch status in the model.

[0031] The present invention establishes a node-branch model through the grid topology data, which can accurately reflect the physical connection relationship between each node (such as substation, load point) and branch (such as transmission line, distribution line) in the grid; the node-branch model can show the transmission path of electric energy in the grid, providing a basis for subsequent power flow analysis and electric energy distribution research; the unified data model enables different types of data (such as topology data, equipment parameters, and operating status data) to be standardized, which is convenient for subsequent data integration and analysis. By integrating key equipment parameters into the node-branch model, the specific parameters of each device (such as transformer capacity, line impedance, circuit breaker rated current, etc.) can be fully understood, which improves the accuracy of the model; the integration of equipment parameters enables the model to reflect not only the topology of the grid, but also the specific operating characteristics and performance of the equipment, which improves the accuracy of the model; comprehensive equipment parameter integration helps the management and maintenance of grid equipment, facilitates the timely discovery and resolution of equipment problems, and improves the operating efficiency and reliability of the grid. By assigning state values ​​to each node and branch through real-time operation status data, the operation status of the power grid, such as voltage, current, power, frequency, etc., can be dynamically reflected. The real-time state value assignment enables the power grid digital twin model to monitor the operation of the power grid in real time, and timely discover and warn potential faults and anomalies. The integration of real-time state data provides accurate decision support for power grid operation management, and can make scientific operation adjustments and optimizations based on real-time data to improve the operation efficiency and safety of the power grid. Combining topological structure data, key equipment parameters and real-time operation status data, the constructed power grid digital twin model can comprehensively and accurately reflect the actual operation of the power grid. Through dynamic state value assignment and real-time monitoring, the model can timely discover anomalies and potential faults in the operation of the power grid, and improve the operation monitoring and early warning capabilities of the power grid. The power grid digital twin model provides a scientific basis for the operation optimization and management of the power grid, and can make operation decisions and optimizations based on real-time data and accurate models to improve the overall efficiency and safety of the power grid. Through the real-time state evaluation and fault early warning of the model, it can support the intelligent operation and maintenance of the power grid and rapid fault handling, reduce the impact of faults and power outage time, and improve the reliability and stability of the power grid. Through the implementation of the above steps, the construction of a digital twin model of the power grid can comprehensively enhance the operation monitoring, management and optimization capabilities of the power grid, and provide strong support for the safe, reliable and efficient operation of the power grid.

[0032] Preferably, step S16 comprises the following steps: Step S161: Calculate node voltage, branch current, active power and reactive power according to the digital twin model of the power grid, and perform system frequency and phase angle difference statistics to obtain a real-time state quantity data set; The embodiment of the present invention uses the digital twin model of the power grid constructed in step S15, which includes the topological structure, equipment parameters and real-time status data of the power grid; the voltage value of each node is calculated using the node voltage calculation formula in the model (such as the calculation method based on the power flow equation) using the real-time power grid data; the current value on the branch is calculated using the node voltage and branch impedance (resistance and reactance). Ohm's law and the power flow equation can be used: ,in, For branch The current, and is the voltage of the two nodes, is the branch impedance; the power flow is calculated using the node voltage and branch current: ,in is the phase angle between current and voltage; the reactive power is calculated using the phase difference between voltage and current: ; Use real-time data or synchronized phasor measurement (PMU) data to determine system frequency; Evaluate the phase angle stability of the grid by calculating the phase angle difference between each node: .

[0033] Step S162: performing static stability analysis based on voltage stability margin on the real-time state quantity data set, thereby obtaining static stability data; The voltage stability margin of the embodiment of the present invention is generally expressed as a critical point of the system voltage, relative to the increase or decrease of the node load, and the calculation method includes using a voltage stability index, such as an L-index or a VQ curve; using a static power flow analysis (such as the Newton-Raphson method) to calculate the voltage stability margin of the power grid under various load conditions, solving the power flow equation, and evaluating whether the voltage is within the allowable range; extracting voltage stability margin data from the power flow analysis results, including the margin of each node voltage; generating a voltage stability report to indicate potential areas of voltage instability.

[0034] Step S163: performing a dynamic stability analysis based on a small disturbance response simulation on the real-time state quantity data set, thereby obtaining dynamic stability data; The embodiment of the present invention uses a small disturbance model to perform dynamic stability analysis, such as a linearized dynamic model or a small disturbance stability model; uses a dynamic simulation tool (such as MATPOWER, PSAT) to simulate the power grid model to simulate the impact of small disturbances on the dynamic behavior of the system. The frequency response, phase angle oscillation and stability index of the system are extracted from the simulation results; and a dynamic stability report is generated, including the response characteristics and stability evaluation of the system under small disturbances.

[0035] Step S164: merging the static stability data and the dynamic stability data into system stability index data; The embodiment of the present invention combines the data in steps S162 and S163 to form a comprehensive system stability index. The data fusion can be performed using weighted average or multivariate analysis methods; a comprehensive system stability index is generated, such as a voltage stability index and a dynamic response stability index; and a comprehensive system stability report is output to describe the overall stability of the power grid, including static and dynamic stability assessments.

[0036] Step S165: Perform safety constraint verification on the power grid through the power grid digital twin model to obtain equipment safety margin data, wherein the safety constraint verification specifically includes verifying whether the node voltage is within the allowable range and checking whether the generator output exceeds the limit; The embodiment of the present invention sets a voltage limit range (such as ±10% rated voltage) to verify whether the voltage of each node is within the allowable range; sets the maximum and minimum output limits of the generator to check whether the actual output of the generator exceeds the limit range. Use the power grid digital twin model to perform safety constraint checks on the power grid to verify whether various safety standards are met; record the safety margin data of each device, including voltage margin and generator output margin.

[0037] Step S166: Perform a system risk quantitative assessment on the equipment safety margin data and the system stability index data, thereby generating fault risk warning data.

[0038] The embodiment of the present invention combines the equipment safety margin data and the system stability index data to calculate the risk level of the system. A risk matrix or a quantitative model (such as probabilistic risk assessment) can be used; based on the safety margin data and stability index, a system risk quantitative model is established to assess the overall risk level of the power grid; based on the risk assessment results, fault risk warning data is generated, including potential fault areas, risk levels, and countermeasures.

[0039] The present invention can comprehensively monitor the operation status of the power grid by calculating node voltage, branch current, active power, reactive power, system frequency and phase angle difference; the acquisition of real-time state quantity data set can provide the latest data basis for subsequent analysis, improve the accuracy and timeliness of analysis; different types of state quantity data (such as voltage, current, power, frequency, etc.) can support multi-dimensional power grid operation analysis and improve monitoring and decision-making capabilities. Static stability analysis can evaluate the voltage stability of the power grid under the current operating state and identify potential voltage stability problems; by analyzing the voltage stability margin, it can warn of voltage stability problems in the power grid in advance to prevent voltage collapse and system instability; according to static stability data, the power grid dispatching strategy can be optimized to ensure that the voltage is within a safe range and improve system stability. Dynamic stability analysis can evaluate the power grid's response capability to small disturbances and identify potential dynamic stability problems; by simulating small disturbance responses, it can prevent system oscillation and instability and improve the dynamic stability of the power grid; dynamic stability data can guide the adjustment of the power grid's control strategy and enhance the system's anti-disturbance and recovery capabilities. The merging of static and dynamic stability data can provide comprehensive system stability indicators and comprehensively evaluate the stability of the power grid; the unification of system stability indicator data is conducive to the formation of a standardized stability evaluation system, which is convenient for comparison and analysis; the comprehensive stability indicator data can provide a comprehensive basis for power grid operation optimization and dispatching decisions, and improve the safety and stability of power grid operation. Through safety constraint verification, it can ensure that the operating parameters of each device in the power grid are within the safe range to prevent equipment overload and damage; the acquisition of equipment safety margin data can help improve the reliability and safety of the power grid system and prevent accidents; the use of digital twin models for safety constraint verification can realize dynamic safety monitoring of the power grid operation status and timely discover and deal with potential safety problems. Through the comprehensive analysis of equipment safety margin and system stability indicators, the fault risk level of the power grid can be quantitatively evaluated; the generation of fault risk warning data can accurately warn high-risk areas and equipment in the power grid and take preventive measures in advance; the quantitative assessment of system risk provides a scientific basis for the risk management of the power grid, which can optimize fault emergency plans and response strategies and improve the emergency handling capabilities of the power grid. Through detailed data collection, status monitoring, stability analysis and risk assessment, the operation monitoring and assessment capabilities of the power grid can be comprehensively improved; real-time status assessment and fault risk warning can prevent potential problems in advance, optimize power grid dispatching strategies, and improve system stability and safety; Quantified risk assessment and accurate warning data will help the power grid to carry out efficient emergency handling when a fault occurs, reduce the impact of faults, and improve power supply reliability; through high-precision digital twin models and comprehensive data analysis, it can support the intelligent operation and maintenance and scientific decision-making of the power grid, and improve the level of intelligence in power grid management.

[0040] Preferably, step S17 comprises the following steps: Step S171: Acquire a fault feature library, which includes feature patterns of various known faults; The embodiment of the present invention collects characteristic patterns of various power grid faults from historical fault records, experimental data, industry standards and technical literature, for example, short circuit faults, equipment overloads, circuit breaker faults, etc.; defines the characteristics of each fault mode, such as current and voltage waveform changes, frequency response, phase angle anomalies, etc.; organizes these fault modes and their characteristics into a database, which can be stored using a relational database or a NoSQL database, including information such as fault type, characteristic description, and occurrence conditions. The feature library is regularly updated from new fault records and the latest research to ensure that it contains the latest fault mode and characteristic information; and a version control tool is used to manage the update records and change history of the feature library.

[0041] Step S172: performing key feature recognition on the fault feature library to obtain key feature data, wherein the key feature recognition specifically refers to extracting operating parameters based on the different fault feature modes; The embodiment of the present invention uses feature selection techniques (such as principal component analysis PCA and linear discriminant analysis LDA) to identify and extract key features of each fault mode. These features may include waveform amplitude, frequency variation, harmonic components, etc.; pattern recognition algorithms (such as K-means clustering and support vector machine SVM) are used to distinguish different fault modes and their key features. The extracted key features are standardized to have a unified measurement unit for subsequent analysis; the extracted key feature data is stored in a fault feature library for subsequent fault mode matching and anomaly detection.

[0042] Step S173: extracting normal characteristics of the power grid from the real-time operation status data, and performing abnormal identification, thereby obtaining abnormal operation data of the power grid; The embodiment of the present invention obtains real-time operation data of the power grid, such as current, voltage, frequency, power, etc., from the real-time monitoring system, and extracts the characteristics of the power grid under normal operation, such as mean, standard deviation, fluctuation range, etc. Anomaly detection algorithms (such as Z-Score, Isolation Forest, Local Anomaly Factor LOF) are used to analyze the real-time data, identify data points that do not conform to normal characteristics, and record abnormal data, including abnormal type, time, location and other information.

[0043] Step S174: extracting key features of abnormal operation data of the power grid according to the key feature data, thereby obtaining real-time operation key feature data; The embodiment of the present invention matches the abnormal operation data of the power grid with the key features in the fault feature library, extracts the key features related to the abnormality, further analyzes the abnormal data using statistical analysis and machine learning algorithms (such as decision trees and random forests), and extracts the most representative key features. The extracted key feature data is organized into a report, including feature values, abnormal situation descriptions, etc., and the key feature data is updated regularly to facilitate subsequent fault pattern matching.

[0044] Step S175: performing fault feature pattern matching on the real-time operation key feature data according to the fault feature library, and confirming the occurrence of the fault according to the preset fault discrimination threshold, thereby obtaining real-time fault condition data, wherein the real-time fault condition data includes the occurrence of the fault and the non-occurrence of the fault.

[0045] The embodiment of the present invention uses a pattern matching algorithm (such as similarity measurement, distance measurement, neural network matching) to match real-time operation key feature data with feature patterns in a fault feature library, calculates the similarity of feature pattern matching, and analyzes whether it meets the characteristics of known fault patterns; sets a fault discrimination threshold based on historical data and expert knowledge, such as a matching similarity threshold, compares the matching result with the threshold, confirms whether a fault occurs, and if the similarity exceeds the threshold, confirms that a fault occurs; records the occurrence of the fault, including the fault type, location, time, etc., and generates a real-time fault condition data report.

[0046] The present invention can accumulate and share characteristic patterns of various known faults by acquiring and maintaining a fault feature library, thereby improving the basis for identifying power grid faults; the fault feature library provides a clear basis and reference for fault identification, which helps to improve the accuracy of fault identification; the fault feature library can be continuously updated and expanded, including new fault modes, to ensure the continuous improvement of fault identification technology. Through key feature identification, the most discerning key features for fault identification can be extracted from the fault feature library, the amount of data can be streamlined, and the processing efficiency can be improved; the key feature data can more effectively distinguish different fault modes, enhance the accuracy and reliability of fault identification; after extracting the key features, the data dimension is reduced, the computational complexity in the fault identification process is reduced, and the real-time response capability is improved. The normal characteristics of the power grid are extracted to provide a benchmark for subsequent abnormal identification and improve the accuracy of abnormal identification; by analyzing the operating status data in real time, the abnormalities in the power grid operation can be identified in time to prevent the further deterioration of the fault; abnormal identification based on real-time data makes power grid monitoring more data-driven, reduces manual intervention, and improves the level of automation. Extracting key features from abnormal data helps focus on the most important parameters and improve the accuracy and efficiency of fault identification; key feature extraction simplifies the data processing process, making subsequent fault identification and pattern matching more efficient; real-time fast extraction of key feature data helps accelerate the fault identification process and improve system response speed. Through fault feature pattern matching, the type of fault in the power grid can be accurately identified to improve the accuracy of fault diagnosis; the preset fault discrimination threshold enables the rapid confirmation of the occurrence of faults, which helps to take emergency measures in a timely manner; real-time fault situation data includes information on the occurrence and non-occurrence of faults, providing a detailed basis for fault handling and recovery of the power grid; through fast and accurate fault identification and confirmation, it helps to improve the stability and security of the power grid and reduce the impact of faults on the system. By building and utilizing a fault feature library, as well as identifying and extracting key features, the accuracy and reliability of power grid fault identification can be greatly improved; real-time extraction of anomalies and key features in operating data, and fault pattern matching, can enhance the real-time monitoring and early warning capabilities of the power grid; rapid confirmation of fault occurrence can help to take effective emergency measures in a timely manner, improve fault handling efficiency, and reduce the impact of faults on the power grid; the implementation of these steps has promoted the intelligent process of power grid fault emergency rescue methods, improved the intelligence and automation level of power grid management, and ensured the safety and stability of power grid operation.

[0047] Preferably, step S2 comprises the following steps: Step S21: when the real-time fault condition data indicates that a fault has occurred, high-resolution fault waveform data before and after the fault occurs is obtained based on the fault risk warning data and the real-time operation status data; Step S22: filtering the high-resolution fault waveform data for noise, and locating the fault point using a pre-trained convolutional neural network, thereby obtaining preliminary fault point location data; Step S23: classifying the high-resolution fault waveform data into fault types based on a support vector machine according to the fault feature library, thereby obtaining fault type identification data; Step S24: fusing the preliminary fault point location data, the fault type identification data and the real-time operation status data to obtain fault diagnosis data; Step S25: performing dynamic response characteristic analysis on the high voltage direct current and flexible alternating current transmission systems in the power grid according to the preset FACTS and HVDC characteristic models and fault diagnosis data, thereby obtaining system response characteristic data; Step S26: performing fault propagation path simulation based on graph theory algorithm according to the system response characteristic data, thereby obtaining fault propagation path prediction data; Step S27: construct a fault impact range model based on the fault propagation path prediction data, the system response characteristic data and the fault diagnosis data.

[0048] As an embodiment of the present invention, refer to Figure 3 As shown, Figure 1 Detailed step flow diagram of step S2 in the embodiment of the present invention, step S2 includes the following steps: Step S21: when the real-time fault condition data indicates that a fault has occurred, high-resolution fault waveform data before and after the fault occurs is obtained based on the fault risk warning data and the real-time operation status data; The embodiment of the present invention uses a real-time monitoring system (such as a SCADA system or other power monitoring system) to collect waveform data of the power grid, including current, voltage, etc.; extracts high-resolution data from the monitoring system, which should have a sufficient sampling frequency (such as 1 kHz or higher) to capture the detailed waveform of the fault. Determine the time point when the fault occurs, and extract the data segments before and after the fault occurs from the data storage system; ensure that the data is saved in an appropriate format (such as CSV, HDF5) to facilitate subsequent processing and analysis.

[0049] Step S22: filtering the high-resolution fault waveform data for noise, and locating the fault point using a pre-trained convolutional neural network, thereby obtaining preliminary fault point location data; The embodiment of the present invention uses digital filtering technology (such as low-pass filter, Kalman filter) to filter noise from high-resolution waveform data, selects appropriate filter parameters to remove background noise and retain signal characteristics; applies smoothing technology (such as moving average method) to further process data to reduce the impact of random noise. Use a pre-trained convolutional neural network model (such as ResNet, VGGNet), which has been trained on similar power fault data and can effectively extract fault features. Input the high-resolution waveform data into the CNN model, pre-process the data (such as normalization) to meet the model input requirements; the CNN model analyzes the waveform data and locates the specific location of the fault, and obtains preliminary fault point location data, including the time and location of the fault.

[0050] Step S23: classifying the high-resolution fault waveform data into fault types based on a support vector machine according to the fault feature library, thereby obtaining fault type identification data; The embodiment of the present invention extracts relevant features from high-resolution fault waveform data, such as waveform amplitude, frequency component, waveform shape, etc. These features will be used for subsequent classification; the extracted features are standardized to ensure data consistency and improve the performance of the classification model. The SVM classification model is trained using labeled fault data (including samples of different fault types). An appropriate kernel function (such as RBF kernel, polynomial kernel) is selected to improve classification accuracy; the extracted features are input into the SVM model for classification, and the model outputs fault type identification data, indicating the specific type of fault (such as short circuit fault, open circuit fault, etc.).

[0051] Step S24: fusing the preliminary fault point location data, the fault type identification data and the real-time operation status data to obtain fault diagnosis data; The embodiment of the present invention uses data fusion algorithms (such as Kalman filtering, weighted average method, data fusion framework) to fuse preliminary fault point location data, fault type identification data and real-time operation status data to ensure that data from different sources can complement each other; the fused data is integrated into a unified format to facilitate subsequent analysis and decision-making, including detailed information, location, type and related operation status of the fault. The fused data is analyzed using a fault diagnosis model (such as a rule engine, decision tree) to confirm the specific nature and impact of the fault and generate the final fault diagnosis data, including the cause of the fault, the scope of impact, etc.

[0052] Step S25: performing dynamic response characteristic analysis on the high voltage direct current and flexible alternating current transmission systems in the power grid according to the preset FACTS and HVDC characteristic models and fault diagnosis data, thereby obtaining system response characteristic data; The embodiment of the present invention establishes a dynamic response characteristic model based on the technical specifications and characteristics of FACTS (Flexible AC Transmission System) and HVDC (High Voltage Direct Current Transmission System) in the power grid, including the control strategy, dynamic behavior, stability characteristics, etc. of the system; the fault diagnosis data is input into the model for analysis, and the impact of the fault on the system is considered, including current, voltage, power fluctuations, etc. Dynamic response characteristics of the power grid system are analyzed using dynamic simulation tools (such as PSCAD, MATLAB / Simulink), and the behavior of the system after the fault occurs is simulated, including oscillation response, stability recovery, etc.; the simulation results are extracted to generate system response characteristic data, including dynamic response curves, stability indicators, recovery time, etc.

[0053] Step S26: performing fault propagation path simulation based on graph theory algorithm according to the system response characteristic data, thereby obtaining fault propagation path prediction data; The embodiment of the present invention abstracts the power grid model into a graph model, in which nodes represent power grid equipment (such as transformers and lines), and edges represent power grid connection relationships. The dynamic characteristics of the graph model are constructed based on the system response characteristic data; graph theory algorithms (such as Dijkstra algorithm and A* algorithm) are used to simulate the path of fault propagation from the fault point, taking into account the impact of the fault on various parts of the system and the propagation rate. Through the simulation results, the possible propagation path of the fault is analyzed, and the expansion range and impact area of ​​the fault on the power grid are predicted; fault propagation path prediction data is generated, including information such as the direction, speed, and impact area of ​​the fault propagation.

[0054] Step S27: construct a fault impact range model based on the fault propagation path prediction data, the system response characteristic data and the fault diagnosis data.

[0055] The embodiment of the present invention combines fault propagation path prediction data, system response characteristic data and fault diagnosis data to establish a fault impact range model. The model should be able to describe the specific impact area of ​​the fault on the power grid, including power equipment, transmission lines and load areas; adjust the model parameters according to historical data and real-time data to ensure that it accurately reflects the actual impact of the fault on the system. Use historical fault cases and actual operation data to verify the accuracy of the model. Compare the impact range predicted by the model with the actual fault impact; optimize the model parameters and structure based on the verification results to improve the prediction ability and reliability of the model.

[0056] The present invention obtains high-resolution fault waveform data, provides accurate raw data for subsequent fault analysis, and improves the accuracy of analysis results; waveform data before and after the fault occurs helps to fully understand the fault process, provide information on fault precursors and fault consequences, and facilitate fault tracing; high-resolution data contains more details and information, and can more accurately reflect the actual situation of the fault. By filtering out noise, the quality of waveform data is improved to ensure the accuracy of subsequent analysis; the pre-trained convolutional neural network is used to automatically locate the fault point, improve the speed and accuracy of fault location, and reduce manual intervention; the efficient processing capability of the convolutional neural network can quickly process large-scale data and improve the efficiency of fault location. Support vector machines are used to classify fault types to improve the accuracy and reliability of fault type identification; classification combined with a fault feature library can make full use of known fault modes and improve the accuracy of identification; the fast classification capability of the support vector machine can identify the fault type in a short time and improve the response speed of fault handling. Through data fusion, data from different sources are comprehensively analyzed to improve the accuracy and comprehensiveness of fault diagnosis; the initial fault point location data, fault type identification data and real-time operation status data are integrated to provide more comprehensive and accurate fault diagnosis results; data fusion can reduce redundant information, improve data processing efficiency, and speed up the fault diagnosis process. Analyze the dynamic response characteristics of high-voltage direct current and flexible alternating current transmission systems under fault conditions to provide important references for fault handling and system recovery; use the preset FACTS and HVDC characteristic models to more accurately simulate the response of the actual system and improve the reliability of the analysis results; dynamic response characteristic data provides a basis for subsequent fault propagation path prediction and impact range modeling, supporting more effective emergency decision-making. Simulate fault propagation paths based on graph theory algorithms to help foresee areas that may be affected by faults and improve the emergency response capabilities of power grids; by predicting fault propagation paths, emergency handling paths can be optimized to reduce the impact range of faults on power grids; understanding fault propagation paths in advance helps prevent and control secondary faults and protect the overall stability of power grids. Construct a fault impact range model to provide a comprehensive assessment of the impact of the fault and help formulate a more accurate emergency response plan; the fault impact range model provides accurate decision-making support for power grid managers to ensure the effectiveness of fault handling measures; by accurately assessing the fault impact range, it helps to improve the grid's recovery and fault resistance capabilities and ensure the safe and stable operation of the grid.The application of technologies such as high-resolution waveform data, noise filtering, deep learning, and support vector machines has significantly improved the accuracy and reliability of fault identification. The automated processing and rapid response capabilities of each step enable the power grid to quickly locate, identify, and diagnose faults when they occur, thereby improving the efficiency of emergency response. Through comprehensive data analysis and modeling, accurate fault diagnosis and impact assessment are provided, providing intelligent decision-making support for power grid managers and improving the scientific nature and effectiveness of emergency response. The combined effects of each step significantly improve the safety and stability of the power grid when faults occur, ensuring that the power grid can quickly resume normal operation and reducing the impact of faults on the power grid and users.

[0057] Preferably, step S3 comprises the following steps: Step S31: determining the equipment and lines directly affected by the fault according to the fault impact range model, thereby obtaining fault impact range data; The embodiment of the present invention uses the fault impact range model to analyze the specific impact of the fault on the power grid equipment and lines, which includes identifying which equipment and lines in the power grid are directly affected by the fault; analyzing the status of the power grid equipment (such as transformers, circuit breakers, switches, etc.) and lines (such as transmission lines, distribution lines) to determine whether these equipment and lines are within the fault impact range. Generate fault impact range data to record which equipment and lines are affected, the degree and scope of the impact; the data format can include information such as the affected area, equipment status, and line load.

[0058] Step S32: using a graph partitioning algorithm to perform regional partitioning on the power grid topology data according to the fault impact range data, thereby obtaining a set of solutions for isolating fault regions; The embodiment of the present invention selects an appropriate graph segmentation algorithm (such as K-means clustering, spectral clustering, graph cut algorithm, etc.) to perform regional segmentation on the power grid topology structure. The graph segmentation algorithm can segment the power grid graph model into multiple regions, and the equipment and lines in each region have similar characteristics; the fault impact range data is used as the input of the graph segmentation algorithm, and the graph segmentation algorithm segments the power grid topology map into multiple regions to isolate the fault area. According to the segmentation results, the boundaries of each region are defined, and the fault impact area and the isolation area are marked. The regional segmentation should take into account the connection relationship and fault propagation characteristics of the power grid; according to the segmentation results, a set of isolation fault area schemes is generated, and each scheme corresponds to a specific regional segmentation scheme, which is used to isolate the fault area and reduce the impact of the fault on the power grid.

[0059] Step S33: Determine the load priority based on the heuristic algorithm according to the real-time operation status data and the power grid topology data, so as to obtain a set of load transfer schemes; The embodiment of the present invention uses heuristic algorithms (such as genetic algorithms, simulated annealing algorithms, particle swarm optimization, etc.) to determine the priority of loads. These algorithms can handle complex optimization problems and find the best load transfer scheme; real-time operating status data (such as load demand, equipment status, operating capacity) and power grid topology data are used as inputs of the heuristic algorithm. According to the operating requirements of the loads, the operating status of the system and the priority requirements, the priority of each load is evaluated, and factors including load importance, power supply reliability and economy are considered; a set of load transfer schemes is generated, and the priority and transfer order of the loads are listed, so as to reasonably transfer and redistribute the loads in the event of a fault.

[0060] Step S34: performing mixed integer programming according to preset power generation resource data and fault impact range data, thereby obtaining a power source reconstruction solution set; The embodiment of the present invention establishes a mixed integer programming model, including power generation resource data (such as generator capacity, generator operating status) and fault impact range data. The model should take into account the start and stop of power generation resources, load demand and system constraints; the mixed integer programming model is solved using an optimization solver (such as Gurobi, CPLEX). This process will generate a set of optimized power reconstruction schemes to meet the needs of the power grid and restore the stability of the system. A power reconstruction scheme set is generated based on the optimization results, including the enabled or disabled generator sets, the scheduling order of the generators, etc. Each scheme should be able to provide effective power support in the event of a fault.

[0061] Step S35: performing adaptive topology network reconstruction on the power grid according to the comprehensive power grid data and the fault impact range data, thereby obtaining a dynamically reconstructed power grid topology solution set; The embodiment of the present invention formulates an adaptive topology reconstruction strategy based on comprehensive grid data (such as load distribution, equipment status) and fault impact range data. The strategy should include line switch operation, equipment start and stop, and grid structure adjustment; perform topology reconstruction operations, adjust the switch status in the grid, and reconfigure line connections to optimize the grid structure and enhance the system's fault recovery capabilities. Generate a set of dynamically reconstructed grid topology solutions, record different grid topology adjustment solutions, and these solutions should be optimized based on real-time data and fault impact range to restore the stability and reliability of the grid.

[0062] Step S36: merge the dynamic reconstruction grid topology solution set, power reconstruction solution set, fault area isolation solution set and load transfer solution set into an emergency sub-solution set; simulate and verify the emergency sub-solution set, and conduct feasibility and impact assessment on each sub-solution, and combine the sub-solutions to obtain optimized emergency solution data.

[0063] The embodiment of the present invention integrates the dynamic reconstruction of the power grid topology solution set, the power reconstruction solution set, the isolation fault area solution set and the load transfer solution set into a comprehensive emergency sub-solution set; each sub-solution represents a complete emergency treatment solution, including topology reconstruction, power supply adjustment, fault isolation and load transfer. Use simulation tools (such as PowerWorld, ETAP) to simulate and verify the emergency sub-solution set, simulate the actual effects and system responses of different emergency solutions under fault conditions; analyze the simulation results and evaluate the feasibility and effect of each solution. Factors of concern include system recovery time, load recovery, equipment safety, etc. Evaluate each sub-solution according to preset standards (such as recovery time, economic cost, equipment safety), and select the solution that best meets the requirements. Combine the various sub-solutions to form the optimal emergency treatment solution, ensure that all emergency measures are coordinated and consistent, and form a comprehensive emergency response strategy. Generate the final optimized emergency solution data, including implementation steps, resource allocation, expected effects, etc., to ensure that the emergency solution can effectively solve the fault problem and restore the normal operation of the power grid.

[0064] The present invention can accurately determine the equipment and lines affected by the fault through the fault impact range model, and provide a more targeted emergency response; accurately identify the affected equipment, avoid secondary faults caused by misoperation, and improve the safety of fault handling; clarify the affected range, which is helpful to quickly formulate emergency handling measures and shorten the fault handling time. The graph partitioning algorithm is used to isolate the power grid in a region, which can effectively isolate the fault area and prevent the fault from spreading; through algorithm optimization, the isolation plan is quickly generated to improve the efficiency of isolating the fault area. The heuristic algorithm is used to intelligently determine the load priority according to the real-time operation status data and the power grid topology data to ensure that the key load is prioritized; through reasonable load priority determination, the optimal load transfer is achieved to reduce the impact of the fault on the power supply; the load transfer plan is quickly generated to shorten the emergency response time and ensure the continuity of the power supply; the fault area is effectively isolated, the impact of the fault on other areas of the power grid is reduced, and the overall stability of the power grid is maintained. The mixed integer programming is used to reconstruct and optimize the power supply to ensure the power supply stability of the power grid under fault conditions; by optimizing the power reconstruction plan, the utilization efficiency of power generation resources is improved and resource waste is reduced; the optimized power reconstruction plan improves the recovery ability of the power grid after the fault and ensures the stability of the power supply. Adaptive topology network reconstruction can dynamically adjust the grid structure according to real-time conditions to ensure the flexibility and reliability of grid operation; by dynamically reconstructing the grid topology, the grid's tolerance and recovery capabilities to faults can be improved, and the impact of faults on the grid can be reduced; the adaptive network reconstruction solution helps optimize the overall operation of the grid and improve the stability and efficiency of the power system. Integrate and comprehensively consider different emergency sub-schemes to provide a more comprehensive emergency response strategy to ensure the comprehensiveness and effectiveness of emergency handling; conduct feasibility assessments on emergency sub-schemes through simulation verification to ensure the feasibility and effectiveness of each scheme in actual operation; combine and optimize each sub-scheme to provide the best emergency handling solution to improve the efficiency and effectiveness of fault handling; the optimized emergency plan data can minimize the impact of faults on the grid and users, and improve the overall reliability and stability of the grid. Through precise analysis and optimization of each step, the accuracy of fault identification, isolation and processing is improved, ensuring the effectiveness of emergency response; using dynamic reconstruction and optimization algorithms, the flexibility and adaptability of the power grid in fault conditions are improved, ensuring that the power grid can quickly resume normal operation; comprehensive analysis and simulation verification provide scientific decision-making support, helping power grid managers to formulate more reasonable and effective emergency response plans; by optimizing emergency plans, the stability and reliability of the power grid are improved, ensuring the continuity and safety of power supply, and reducing the impact of faults on users.

[0065] Preferably, step S361 includes the following steps: Step S361: merging the dynamic power grid topology reconstruction solution set, the power source reconstruction solution set, the fault area isolation solution set and the load transfer solution set into an emergency sub-solution set; The embodiment of the present invention collects all the solutions in the set of dynamic reconstruction of power grid topology solutions, power reconstruction solutions, isolation fault area solutions and load transfer solutions, which involve power grid recovery strategies at different levels; assigns a unique identifier to each solution, and records relevant information (such as execution steps, resource requirements, expected effects, etc.). The above solution sets are integrated into a comprehensive emergency sub-solution set using data fusion technology (such as set operations and matching algorithms), and each emergency sub-solution includes strategies for power grid topology adjustment, power supply configuration, fault isolation and load transfer; ensure that the merged emergency sub-solution set is complete, covers all possible emergency treatment measures, and each sub-solution has reasonable implementation steps.

[0066] Step S362: Rapidly simulate each scheme of the emergency sub-scheme set according to the power grid digital twin model by using parallel computing technology, so as to obtain a simulation result data set; The embodiment of the present invention configures a parallel computing environment (such as a multi-core processor, a cluster computing platform, or cloud computing resources) to ensure that computing resources can handle multiple simulation tasks at the same time; uses a parallel computing framework (such as MPI, OpenMP, Hadoop, etc.) to allocate and manage computing tasks, and allocates each emergency sub-plan to different computing nodes for simulation. Load the corresponding power grid status and parameters according to the power grid digital twin model, and apply each emergency sub-plan for simulation; execute simulation tasks to simulate the implementation effect of each emergency sub-plan. Including power grid recovery, load transfer, power supply scheduling, etc. Collect the result data of each simulation task, including system recovery time, equipment load, fault handling effect, etc.; summarize all simulation results to generate a simulation result data set for subsequent evaluation and optimization.

[0067] Step S363: Perform feasibility and impact assessment based on system recovery time requirements, cost constraints, and safety constraints according to the simulation result data set, and establish multiple objective optimization models to obtain an emergency multi-objective optimization model; The embodiment of the present invention determines the time requirement for system recovery, such as the shortest recovery time or the maximum tolerable recovery time, sets cost constraints, including operating costs, equipment damage repair costs, etc., and determines safety constraints, such as equipment operation safety, grid stability, etc. The feasibility and impact of each sub-scheme are evaluated using the data in the simulation result data set. The evaluation indicators may include system recovery time, cost, equipment load, safety, etc. An evaluation score is assigned to each sub-scheme to indicate its performance in meeting the recovery time, cost, and safety constraints. A multi-objective optimization model is established based on the evaluation results. The model should consider multiple objectives (such as minimum recovery time, minimum cost, maximum safety) and set corresponding constraints; the objective function of the optimization model is defined, which usually includes a weighted combination of various evaluation indicators.

[0068] Step S364: solving the multi-objective optimization problem of the emergency multi-objective optimization model by using a non-dominated sorting genetic algorithm, thereby obtaining a Pareto optimal solution set; The embodiment of the present invention selects the non-dominated sorting genetic algorithm (NSGA-II) as the optimization algorithm, which is suitable for processing multi-objective optimization problems, and configures the parameters of the genetic algorithm, including population size, crossover rate, mutation rate, number of iterations, etc. Generate an initial population, evaluate and sort each individual, perform selection, crossover and mutation operations, generate a new population, and repeat the process until the stop condition is reached; according to the non-dominated sorting, update the population and obtain the Pareto frontier, indicating that there is no possibility of any improvement between the various objectives. Extract the Pareto optimal solution set from the optimization results, indicating the optimal solution combination between the various objectives.

[0069] Step S365: combining the emergency sub-plan set into sub-plans according to the Pareto optimal solution set, thereby obtaining optimized emergency plan data.

[0070] The embodiment of the present invention analyzes the Pareto optimal solution set to understand the performance of each solution on different objectives, and combines multiple sub-plans based on the information in the optimal solution to achieve the best overall effect. This may include selecting partial strategies of multiple sub-plans to form a comprehensive optimized emergency plan; integrating the combined sub-plans into the final optimized emergency plan data. Ensure that each plan can effectively respond to failures and meet system recovery requirements; verify the optimized emergency plan to ensure its feasibility and effectiveness, and if necessary, make further adjustments and optimizations to meet actual operating conditions; output the final optimized emergency plan data, including implementation steps, resource allocation, expected results, etc., to ensure that the plan can provide the best recovery effect under actual failure conditions.

[0071] The present invention integrates multiple emergency plans to form an emergency sub-plan set, provides comprehensive emergency response options, and ensures the comprehensiveness and coordination of the processing process; integrates multiple plan sets to reduce duplication between plans and improve the overall efficiency of emergency response; the emergency sub-plan set provides multiple emergency processing paths and methods to enhance the flexible response capability of the power grid in the event of a fault. By using parallel computing technology, the simulation speed is greatly improved, and multiple emergency plans can be simulated and verified in a short time; parallel computing optimizes resource utilization, and can simultaneously process simulation tasks of multiple emergency plans, improving computing efficiency; through the power grid digital twin model, the accuracy of the simulation results is ensured, and reliable data is provided for subsequent decision-making. Combined with the requirements of system recovery time, cost, safety and other aspects, the feasibility and effectiveness of each emergency plan are evaluated; multiple target optimization models are established, and various constraints and goals are comprehensively considered to provide the optimal emergency treatment plan; the establishment of a multi-target optimization model improves the scientificity and rationality of decision-making, and ensures the feasibility and efficiency of the emergency plan in actual operation. The use of non-dominated sorting genetic algorithm can efficiently solve complex multi-objective optimization problems and quickly find the optimal solution set; the Pareto optimal solution set can find the best balance point between multiple objectives, ensuring the optimality of the emergency plan in multiple dimensions; the Pareto optimal solution set provides a variety of optimal solutions, increasing the flexibility and optionality of emergency handling. According to the Pareto optimal solution set, the emergency sub-plans are optimized and combined to ensure that the final emergency plan has the best performance in multiple objectives; the optimized emergency plan is more operational and executable, improving the emergency response efficiency of the power grid in the event of a fault; the final optimized emergency plan data covers all aspects of power grid recovery, ensuring the comprehensiveness and effectiveness of the emergency response. Parallel computing technology and fast simulation have greatly improved the verification efficiency of emergency plans, ensuring reliable simulation results in the shortest time; through multi-objective optimization and the solution of Pareto optimal solution sets, scientific and reasonable emergency decision-making support is provided to ensure the optimality and efficiency of emergency plans; comprehensive integration of multiple emergency plans and optimized combination of final emergency plan data improve the recovery capacity and stability of the power grid after a fault; fast simulation and efficient solution technology ensure that after a fault occurs, an emergency treatment plan can be quickly formulated and implemented to reduce the impact of the fault on the power grid and users.

[0072] Preferably, step S41 includes the following steps: Step S41: obtaining black start capability data of each generator set in the power grid; performing power supply availability and capacity evaluation on the black start capability data, thereby obtaining a black start resource list; The embodiment of the present invention collects the basic information of all generator sets in the power grid, including type (such as coal-fired, gas-fired, nuclear power, wind power, solar power, etc.), model, power generation capacity, operating status and maintenance records; obtains the black start capability data of each generator set, including the start time, start sequence requirements, load bearing capacity, and voltage and frequency response at start-up. Use historical data to analyze the availability of each generator set, including its failure rate, maintenance records, backup capacity, etc.; evaluate the current status of the generator set to confirm whether it can be black started, considering the health status of the equipment, the operating environment, and possible limitations. Calculate the maximum power output that each generator set can provide in the black start state, considering the starting capacity of the equipment, the load demand, and the actual load of the power grid; generate a black start resource list based on the power supply availability and capacity assessment. The list should include the ID, type, maximum output power, start time and other relevant information of the generator set.

[0073] Step S42: constructing a black start capability model according to the black start resource list and corresponding black start performance information; The embodiment of the present invention defines the structure of the black start capability model based on the data in the black start resource list. The model should include the starting capability, response time, load bearing capacity, etc. of the generator set; the performance information of each generator set is integrated into the model, including starting voltage, frequency response, load growth curve, etc. The starting characteristic parameters of the model are set, such as minimum starting load, starting time, starting sequence, etc.; the dynamic characteristics that the generator set may experience during the starting process are defined, including voltage fluctuations, frequency changes, etc. The model is verified using historical data and actual operating data, and the performance of the model under different fault conditions is tested to ensure that it can accurately simulate the actual black start process; according to the test results, the model parameters are adjusted and optimized to ensure the accuracy and practicality of the model.

[0074] Step S43: identifying potential recovery areas for the black start capability model, thereby obtaining black start recovery area data; The embodiment of the present invention divides the potential recovery areas of the power grid according to the black start capability model. These areas generally refer to the parts of the power grid that can be restored by specific generator sets; analyzes the influence range of each generator set, including the area that can be restored, the load distribution and the power grid topology. Generates black start recovery area data, including the specific location of each recovery area, the equipment and load involved, the recovery capacity and other information; assigns a priority to each recovery area according to the capability and influence range of the generator set, and gives priority to restoring the areas that are most critical to the power grid system. Use simulation or actual data to verify the effect of the recovery area to ensure that the division of the recovery area can effectively support the recovery of the power grid; according to the verification results, adjust the division and priority of the recovery area to ensure that it adapts to different fault scenarios and recovery requirements.

[0075] Step S44: Taking the existing resources in the black start resource list as the starting point and the black start recovery area data as the impact range, a recovery strategy for the power grid system is formulated through a heuristic algorithm, thereby generating black start path planning data.

[0076] The embodiment of the present invention matches appropriate generator sets and recovery areas according to the existing resources in the black start resource list and the black start recovery area data, ensuring that each recovery area can get sufficient support. Select a suitable heuristic algorithm (such as genetic algorithm, ant colony algorithm, simulated annealing, etc.) to optimize the recovery strategy; use the heuristic algorithm to generate the recovery strategy of the power grid system, including the start-up sequence of the generator sets, load distribution, power grid topology adjustment, etc. Optimize the black start path planning through the heuristic algorithm to ensure that the power grid system can effectively utilize resources during the recovery process, shorten the recovery time, and meet the safety and stability requirements; verify the generated black start path planning data to test its effectiveness and feasibility in actual fault scenarios. Generate black start path planning data, including detailed steps of the recovery strategy, required resources, expected recovery time, etc.; prepare an implementation plan to ensure that the power grid system can be quickly and effectively restored when an actual fault occurs.

[0077] The present invention systematically obtains the black start capability data of each generator set to ensure that all potential starting resources are considered; by evaluating the power supply availability and capacity, the actually available black start resources are screened out to improve resource utilization; a black start resource list is formed to clarify which resources can be used during the recovery process, laying the foundation for the subsequent recovery strategy formulation. Integrating black start resources and performance information into a model makes resource management more systematic and scientific; through the black start capability model, the performance of each resource can be accurately evaluated to ensure the selection of the best resource combination during the recovery process; the modeling approach makes the management and use of black start resources more efficient and reduces errors in human judgment. Through the analysis of the black start capability model, potential recovery areas are identified to ensure the pertinence and effectiveness of the recovery strategy; identifying potential recovery areas helps to optimize the allocation of black start resources, making the recovery process more efficient and orderly; the generated black start recovery area data provides important data support for the formulation of recovery paths and strategies, ensuring the scientificity and accuracy of the recovery process. Starting from existing resources, ensure that the recovery process starts with reliable resources to improve the success rate of recovery; use heuristic algorithms, combined with black start recovery area data, formulate scientific and reasonable recovery strategies to ensure the optimality of the recovery path; heuristic algorithms can quickly generate recovery path planning data, improve the ability of the power grid to quickly recover after a fault, and reduce the time and scope of power outages; the generated black start path planning data provides a clear roadmap for the recovery of the power grid system, ensuring the systematic and efficient recovery process. Through the establishment of a black start resource list and capability model, comprehensive data support and scientific evaluation are provided for the power grid recovery strategy; reasonable resource evaluation and allocation optimize the utilization of black start resources to ensure that each resource can maximize its utility; the identification of potential recovery areas and the application of heuristic algorithms make the formulation of power grid recovery strategies more scientific and reasonable, and improve the quality of decision-making; scientific and reasonable black start path planning data improves the recovery speed and efficiency of the power grid system after a fault, and reduces the impact of the fault on users and the system.

[0078] Preferably, step S51 includes the following steps: Step S51: Obtain historical load data and user type data, analyze the time-varying characteristics and temperature sensitivity of different types of loads according to the historical load data and user type data, and evaluate the power factor and harmonic characteristics of the load, so as to obtain load characteristic data; The embodiment of the present invention collects historical load data of the power grid, including load demand, power usage pattern and load fluctuation in each time period; collects user type data, such as household users, commercial users, industrial users, etc., and their load characteristics and electricity usage habits. Draw a load curve based on historical load data to identify the time-varying pattern of load demand; calculate the statistical characteristics of load fluctuations, such as peak value, valley value and mean value, and analyze the law of load change over time. Collect temperature data related to the load, such as air temperature, room temperature, etc.; use regression analysis or correlation analysis to evaluate the sensitivity of the load to temperature changes and determine the degree of impact of temperature changes on the load. Calculate the power factor based on voltage and current data and evaluate the power factor of each load; use spectrum analysis technology to perform harmonic analysis on the load current and evaluate the degree of harmonic distortion; collect the harmonic data of the load and analyze its impact on the power grid. Summarize the analysis results to generate load characteristic data, including time-varying characteristics, temperature sensitivity, power factor and harmonic characteristics; record the load characteristic data in a database for subsequent analysis and decision-making.

[0079] Step S52: subdividing the load characteristic data into load categories according to the black start path planning data, and evaluating the load recovery priority of each load category, thereby obtaining a load recovery priority list; The embodiment of the present invention identifies the load areas and categories that need to be restored first based on the black start path planning data; classifies the load data according to functional requirements, importance and restoration priority, such as basic load, critical load and adjustable load. An impact assessment is performed on each type of load, considering its impact on the stability of the power grid system, user needs and restoration time, and a priority score is assigned to each type of load based on the impact analysis results of the load category; the load categories and corresponding priority scores are organized into a load restoration priority list.

[0080] Step S53: Formulate a load recovery strategy according to the load recovery priority list, thereby obtaining a preliminary load recovery strategy; The embodiment of the present invention determines the objectives of the load recovery strategy, including recovery time, load recovery order, and resource utilization efficiency; formulates the order of load recovery based on the load recovery priority list to ensure that the most critical loads to the power grid system are restored first; formulates a detailed load recovery strategy, including load access order, load adjustment, and power grid support measures, in combination with load category characteristics and recovery priorities. The formulated load recovery strategy is recorded and organized into a document, including the implementation steps, expected results, and implementation requirements of the strategy; and the preliminary strategy is reviewed and adjusted to ensure its feasibility and effectiveness.

[0081] Step S54: Perform real-time load monitoring on the load recovery of the power grid system according to the preliminary load recovery strategy, and use an adaptive control algorithm to adjust the load access sequence and power in real time, thereby obtaining load recovery control data.

[0082] The embodiment of the present invention deploys a real-time load monitoring system to collect load data, including actual load, load change rate and grid status; collects load data in real time to monitor dynamic changes and system responses during load recovery. Select appropriate adaptive control algorithms, such as fuzzy control, adaptive PID control, etc., to adjust the load access sequence and power; adjust the load access sequence and power in real time according to real-time load monitoring data and preliminary load recovery strategies to ensure system stability and recovery efficiency; record control data generated during load recovery, including adjusted load access sequence, power changes and system responses; analyze load recovery control data to evaluate the effectiveness of the recovery strategy and system performance. Optimize and adjust the load recovery strategy based on monitoring and control data to ensure the stability and efficiency of the recovery process.

[0083] The present invention can fully understand the time-varying characteristics and temperature sensitivity of different loads by analyzing historical load data and user type data. This helps to predict and understand the behavior of loads under different conditions and optimize the load management of the power grid; evaluating the power factor and harmonic characteristics of the load can reveal the impact of the load on the power quality. This helps to reasonably adjust the load during the recovery process to reduce the negative impact on the power grid; the generated load characteristic data provides a scientific basis for subsequent load classification and recovery strategies, and enhances the accuracy and effectiveness of decision-making. According to the load characteristic data, the load category is subdivided, and the load can be classified according to its characteristics and needs, so that the recovery strategy is more targeted; by evaluating the recovery priority of each type of load, it can be determined which loads should be restored first to minimize the system recovery time and impact; the load recovery priority list helps to optimize resource allocation during the recovery process, making the recovery process more efficient and critical loads can be given priority. The recovery strategy formulated according to the load recovery priority list can ensure that important loads are given priority during the recovery process, improve the overall stability and recovery efficiency of the system; through the priority strategy, load recovery can be better coordinated to reduce possible conflicts and resource waste during the recovery process; a clear load recovery strategy helps to restore power grid services faster and reduce power outage time and inconvenience for users. Real-time load monitoring ensures that every step of the recovery process is being tracked continuously, and the adaptive control algorithm can adjust the load access sequence and power according to the actual situation, making the recovery process more flexible and efficient; the adaptive control algorithm can respond to changes in the load recovery process in real time, adjust the strategy to adapt to the actual operating conditions, and optimize the recovery effect; by adjusting the load access sequence and power in real time, it can reduce the fluctuations that may occur in the system during the recovery process and improve the stability and security of the system. Through detailed analysis and priority evaluation of load characteristics, the formulated recovery strategy can significantly improve the recovery efficiency of the power grid system; load category segmentation and priority evaluation help optimize the use of recovery resources, ensure that key loads are restored first, and improve the overall recovery effect; the application of real-time monitoring and adaptive control algorithms improves the flexibility and system stability of the recovery process and reduces potential problems in the recovery process; clear recovery strategies and effective adjustment mechanisms reduce power outage time and impact on users, and improve user satisfaction.

[0084] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0085] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A power grid fault emergency rescue method, characterized in that: The following steps are involved: Step S1: Obtaining grid topology data, key equipment parameters, and real-time operating status data; Build a digital twin model of the power grid based on the power grid topology data, key equipment parameters, and real-time operating status data, and conduct real-time status assessment to obtain fault risk warning data and real-time fault status data; Step S2: When the real-time fault condition data indicates that a fault has occurred, the fault point is located and the fault type is determined based on deep learning according to the fault risk warning data and the real-time operation status data, thereby obtaining fault diagnosis data; Construct a fault impact range model based on fault diagnosis data and preset FACTS and HVDC characteristic models; Step S3: Generate multiple sets of emergency processing sub-plans based on the multi-objective optimization algorithm according to the fault impact range model to obtain an emergency sub-plan set, perform simulation verification on the emergency sub-plan set, conduct feasibility and impact assessment on each sub-plan, and combine the sub-plans to obtain optimized emergency plan data; Step S4: obtaining a black start capability model of the current power grid, formulating a restoration strategy for the power grid system according to the optimized emergency plan data and the black start capability model, thereby generating black start path planning data; Step S5: Formulate a load recovery strategy based on load characteristics according to the black start path planning data, and use an adaptive control algorithm to adjust the load access sequence and power in real time, thereby obtaining load recovery control data; Step S6: Design a multi-level control system architecture based on the load recovery control data, black start path planning data, and optimized emergency plan data, and use edge computing technology to control the device nodes to obtain a system recovery execution plan.

2. The power grid failure emergency rescue method according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: collecting substations, transmission lines, distribution networks and their physical connection relationships in the power grid, and integrating the devices and their connection relationships to obtain power grid topology data; Step S12: collecting equipment parameters of transformers, transmission lines, circuit breakers, capacitors and reactors, distribution transformer battery energy storage systems, and generators to obtain key equipment parameters; Step S13: Real-time monitoring of voltage, current, power, frequency and power flow information is performed through the substation auxiliary service system to obtain real-time operation status data; Step S14: standardizing the grid topology data, key equipment parameters and real-time operation status data, and removing data with abnormal formats, thereby obtaining comprehensive grid data; Step S15: constructing a power grid digital twin model based on comprehensive power grid data; Step S16: Performing real-time status evaluation on the power grid digital twin model to obtain fault risk warning data; Step S17: Perform feature comparison and abnormality identification based on a preset fault feature library and real-time operating status data, thereby obtaining real-time fault condition data.

3. The power grid failure emergency rescue method according to claim 2, characterized in that: Step S15 includes the following steps: A digital twin model of the power grid is constructed based on the comprehensive data of the power grid. The model construction specifically includes establishing a node-branch model representing the network structure of the power grid and the flow transmission path according to the power grid topology data, adding device nodes and integrating parameters to the node-branch model according to key equipment parameters, and assigning state quantities to each node and branch according to real-time operating status data.

4. The power grid failure emergency rescue method according to claim 3, characterized in that: Step S16 includes the following steps: Step S161: Calculate node voltage, branch current, active power and reactive power according to the digital twin model of the power grid, and perform system frequency and phase angle difference statistics to obtain a real-time state quantity data set; Step S162: performing static stability analysis based on voltage stability margin on the real-time state quantity data set, thereby obtaining static stability data; Step S163: performing a dynamic stability analysis based on a small disturbance response simulation on the real-time state quantity data set, thereby obtaining dynamic stability data; Step S164: merging the static stability data and the dynamic stability data into system stability index data; Step S165: Perform safety constraint verification on the power grid through the power grid digital twin model to obtain equipment safety margin data, wherein the safety constraint verification specifically includes verifying whether the node voltage is within the allowable range and checking whether the generator output exceeds the limit; Step S166: Perform a system risk quantitative assessment on the equipment safety margin data and the system stability index data, thereby generating fault risk warning data.

5. The power grid failure emergency rescue method according to claim 4, characterized in that: Step S17 includes the following steps: Step S171: Acquire a fault feature library, which includes feature patterns of various known faults; Step S172: performing key feature identification on the fault feature library to obtain key feature data, wherein the key feature identification specifically includes extracting operating parameters for distinguishing each fault feature mode; Step S173: extracting normal characteristics of the power grid from the real-time operation status data, and performing abnormal identification, thereby obtaining abnormal operation data of the power grid; Step S174: extracting key features of abnormal operation data of the power grid according to the key feature data, thereby obtaining real-time operation key feature data; Step S175: performing fault feature pattern matching on the real-time operation key feature data according to the fault feature library, and confirming the occurrence of the fault according to the preset fault discrimination threshold, thereby obtaining real-time fault condition data, wherein the real-time fault condition data includes the occurrence of the fault and the non-occurrence of the fault.

6. The power grid failure emergency rescue method according to claim 5, characterized in that: Step S2 includes the following steps: Step S21: when the real-time fault condition data indicates that a fault has occurred, high-resolution fault waveform data before and after the fault occurs is obtained based on the fault risk warning data and the real-time operation status data; Step S22: filtering the high-resolution fault waveform data for noise, and locating the fault point using a pre-trained convolutional neural network, thereby obtaining preliminary fault point location data; Step S23: classifying the high-resolution fault waveform data into fault types based on a support vector machine according to the fault feature library, thereby obtaining fault type identification data; Step S24: fusing the preliminary fault point location data, the fault type identification data and the real-time operation status data to obtain fault diagnosis data; Step S25: performing dynamic response characteristic analysis on the high voltage direct current and flexible alternating current transmission systems in the power grid according to the preset FACTS and HVDC characteristic models and fault diagnosis data, thereby obtaining system response characteristic data; Step S26: performing fault propagation path simulation based on graph theory algorithm according to the system response characteristic data, thereby obtaining fault propagation path prediction data; Step S27: construct a fault impact range model based on the fault propagation path prediction data, the system response characteristic data and the fault diagnosis data.

7. The power grid failure emergency rescue method according to claim 6, characterized in that: Step S3 includes the following steps: Step S31: determining the equipment and lines directly affected by the fault according to the fault impact range model, thereby obtaining fault impact range data; Step S32: using a graph partitioning algorithm to perform regional partitioning on the power grid topology data according to the fault impact range data, thereby obtaining a set of solutions for isolating fault regions; Step S33: Determine the load priority based on the heuristic algorithm according to the real-time operation status data and the power grid topology data, so as to obtain a set of load transfer schemes; Step S34: performing mixed integer programming according to the preset power generation resource data and the fault impact range data, so as to obtain a power source reconstruction solution set; Step S35: performing adaptive topology network reconstruction on the power grid according to the comprehensive power grid data and the fault impact range data, thereby obtaining a dynamically reconstructed power grid topology solution set; Step S36: merge the dynamic reconstruction grid topology solution set, power reconstruction solution set, fault area isolation solution set and load transfer solution set into an emergency sub-solution set; simulate and verify the emergency sub-solution set, and conduct feasibility and impact assessment on each sub-solution, and combine the sub-solutions to obtain optimized emergency solution data.

8. The power grid failure emergency rescue method according to claim 7, characterized in that: Step S361 includes the following steps: Step S361: merging the dynamic power grid topology reconstruction solution set, the power source reconstruction solution set, the fault area isolation solution set and the load transfer solution set into an emergency sub-solution set; Step S362: Rapidly simulate each scheme of the emergency sub-scheme set according to the power grid digital twin model by using parallel computing technology, so as to obtain a simulation result data set; Step S363: Perform feasibility and impact assessment based on system recovery time requirements, cost constraints, and safety constraints according to the simulation result data set, and establish multiple objective optimization models to obtain an emergency multi-objective optimization model; Step S364: solving the multi-objective optimization problem of the emergency multi-objective optimization model by using a non-dominated sorting genetic algorithm, thereby obtaining a Pareto optimal solution set; Step S365: combining the emergency sub-plan set into sub-plans according to the Pareto optimal solution set, thereby obtaining optimized emergency plan data.

9. The power grid failure emergency rescue method according to claim 8, characterized in that: Step S41 includes the following steps: Step S41: obtaining black start capability data of each generator set in the power grid; performing power supply availability and capacity evaluation on the black start capability data, thereby obtaining a black start resource list; Step S42: constructing a black start capability model according to the black start resource list and corresponding black start performance information; Step S43: identifying potential recovery areas for the black start capability model, thereby obtaining black start recovery area data; Step S44: Taking the existing resources in the black start resource list as the starting point and the black start recovery area data as the impact range, a recovery strategy for the power grid system is formulated through a heuristic algorithm, thereby generating black start path planning data.

10. The power grid failure emergency rescue method according to claim 9, characterized in that: Step S51 includes the following steps: Step S51: Obtain historical load data and user type data, analyze the time-varying characteristics and temperature sensitivity of different types of loads according to the historical load data and user type data, and evaluate the power factor and harmonic characteristics of the load, so as to obtain load characteristic data; Step S52: subdividing the load characteristic data into load categories according to the black start path planning data, and evaluating the load recovery priority of each load category, thereby obtaining a load recovery priority list; Step S53: Formulate a load recovery strategy according to the load recovery priority list, thereby obtaining a preliminary load recovery strategy; Step S54: Perform real-time load monitoring on the load recovery of the power grid system according to the preliminary load recovery strategy, and use an adaptive control algorithm to adjust the load access sequence and power in real time, thereby obtaining load recovery control data.

Images